i

The Retail Food Environment: A Rural Perspective

Thesis Submitted By

Rebecca R Harris

Under Supervision of

Dr. Catherine Mah & Dr. James Valcour

In Partial Fulfillment for the Degree of

Master of Science in Medicine (Community Health and Humanities)

Memorial University of Newfoundland

October, 2018

© 2018 Rebecca Harris

ii

ABSTRACT

Through a series of manuscripts, this thesis has critically reviewed, analyzed and

discussed the contextual factors that are significant to measuring and intervening in the

rural retail food environment.

In order to encourage action and influence change, it is most effective to produce

evidence that is meaningful to policymakers, retail owners, and other key stakeholders,

guiding them in prioritized and informed decision making. To do this, we need to

promote greater transparency in reporting methodology; providing explicit definitions and

rationales so that findings are more accessible to knowledge users and can be used to

guide future research and policy direction. Reporting on all aspects of the food

environment including availability, quality and price as well as the exposure to nutritious

and non-nutritious foods are critical to capturing barriers and potential areas for

intervention in rural communities.

Improving food access in rural communities will require a comprehensive and multi-

sectoral effort. A direct and prioritized approach that addresses the fundamental barriers

to food access in rural communities is essential to making impactful improvements in the

accessibility of nutritious foods in rural communities and the health and well-being of

those who live there.

iii

ACKNOWLEDGEMENTS

This research would not have been possible without the time and dedication put forward

by those involved in the Healthy Corner Stores, NL project. I would like to thank the

entire research team for their efforts and for allowing me to be a part of such an

influential project.

I would like to extend my greatest thanks to my supervisor Dr. Catherine Mah. Your

mentorship, support, guidance, kindness and infinite patience have been immeasurably

appreciated. Your success has, and will continue to, provide me with the inspiration to

reach my academic and professional ambitions. I would also like to thank Dr. James

Valcour and Dr. Leia Minaker for their time and resources committed to the completion

of this project.

I would like to thank my peers, classmates and friends for their open ears and open arms

and for giving me the encouragement, confidence and optimism I needed to complete my

thesis. Thank you to my partner Zach, for his continued support and understanding

throughout this endeavor.

Finally, I owe endless thanks to my parents; without you this opportunity and those that

have come from it would not have been possible and for that I am forever thankful.

iv

TABLE OF CONTENTS

Abstract................................................................................................................................ii

Acknowledgements.............................................................................................................iii List of Tables......................................................................................................................vii List of Figures....................................................................................................................vii List of Appendices............................................................................................................viii List of Abbreviations........................................................................................................viii SECTION I: INTRODUCTION CHAPTER 1: Issue History & Theoretical Frameworks Influencing Obesity....................2 1.1 Issue History.......................................................................................................3 1.2 Theoretical Frameworks Influencing Obesity....................................................4 CHAPTER 2: Existing Evidence & Local Context..............................................................9 2.1 Existing Evidence.............................................................................................10 2.2 Local Context...................................................................................................13 CHAPTER 3: Evidence to Action: Measuring the Food Environment and an Introduction to Nutrition Environment Measures Survey.......................................................................15 3.1 Measuring the Food Environment....................................................................16 3.2 The Nutrition Environment Measures Survey..................................................18 CHAPTER 4: Research Rationale, Objectives & Design..................................................23

4.1 Rationale...........................................................................................................24 4.2 Objectives.........................................................................................................26 4.3 Design...............................................................................................................26 4.4 Ethics................................................................................................................27 References..............................................................................................................28

v

SECTION II: MANUSCRIPT SERIES CHAPTER 5: The Nutrition Environment Measures Survey in Rural Communities: A Scoping Review ............................................................................................................32 5.1 Abstract............................................................................................................35 5.2 Introduction......................................................................................................36 5.3 Methods............................................................................................................38 5.4 Results..............................................................................................................41 5.5 Discussion........................................................................................................47 5.6 Conclusion and Recommendations..................................................................53 References..............................................................................................................55 CHAPTER 6: The Nutrition Environment Measures Survey Newfoundland and Labrador: An Exploratory Analysis into the Significance of Assessing the Retail Food Environment Using Healthy Food Only..................................................................................................60 6.1 Abstract............................................................................................................62 6.2 Background......................................................................................................63 6.3 Objectives & Research Design.........................................................................67 6.4 Methods............................................................................................................68 6.5 Results..............................................................................................................72 6.6 Discussion........................................................................................................74 6.7 Strengths and Limitations.................................................................................79 6.8 Conclusion........................................................................................................79 References..............................................................................................................81 Appendices.............................................................................................................85 CHAPTER 7: Fostering Health Retail Food Environments in Rural Newfoundland and Labrador: A Policy Analysis .............................................................................................92 7.1 Abstract............................................................................................................94 7.2 Introduction and Background...........................................................................95 7.3 Fiscal Policy in the Food Environment..........................................................100 7.4 Local Policy....................................................................................................104 7.5 Policy Alternatives.........................................................................................108 7.6 Policy Analysis...............................................................................................113 7.7 Summary & Recommendations for Future Action.........................................128 References............................................................................................................130

vi

SECTION III: SUMMARY & CONCLUSION CHAPTER 8: Summary and Recommendations for Future Research & Policy Direction...........................................................................................................................138 8.1 Introduction....................................................................................................139 8.2 Key Areas of Investigation.............................................................................139 8.3 Key Learnings................................................................................................141 8.4 Recommendations for Future Research and Policy Direction.......................145 8.5 Conclusion......................................................................................................157 References............................................................................................................158

vii

LIST OF TABLES

Table 5-1 Summary of Review Findings.............................................................................46

Table 6-1 Mean Scores by Remoteness..............................................................................73

Table 6-2. Mean Scores by Store Type................................................................................73

Table 6A-1 NL Remoteness Index......................................................................................85

Table 6A-2 North American Industry Classification System for Store Types....................85

Table 6B-1 Healthy Food Availability Score.....................................................................86

Table 6B-2 Unhealthy Food Availability Score.................................................................88

Table 6C-1 Healthy Individual Measure Scores by Store Type.........................................89 Table 7-1 Summary of Policy Considerations...................................................................127

LIST OF FIGURES

Figure 1-1 Ecological Framework of Multifactorial Influences on Diet..............................5

Figure 1-2 The Relationship Among Individual, Environmental and Social Factors............7 Figure 1-3 Model of Community Nutrition Environment.....................................................8 Figure 5-1 Literature Selection...........................................................................................39

Figure 7-1 The Framework for Analysis of Proposed Policies..........................................111

viii

LIST OF APPENDICES

Appendix 6A.......................................................................................................................85

Appendix 6B.......................................................................................................................86

Appendix 6C.......................................................................................................................89

LIST OF ABBREVIATIONS

CCHS Canadian Community Health Survey

HCSNL Healthy Corner Store Newfoundland

HFAI Healthy Food Availability Index

HFFI Healthy Food Financing Initiative

NAICS North American Industry Classification System

NCC Nutrition North Canada

NEMS Nutrition Environment Measures Survey

NEMS-CS Nutrition Environment Measures Survey for Corner Stores

NEMS-NL Nutrition Environment Measures Survey Newfoundland and Labrador

NEMS-S Nutrition Environment Measures Survey for Stores

NL Newfoundland and Labrador

SSBs Sugar Sweetened Beverages

TCII NL Department of Tourism, Culture, Industry and Innovation

ToNEMS Toronto Nutrition Environment Measures Survey

1

SECTION I

Background, Rationale & Objectives

2

CHAPTER 1: Issue History & Theoretical Frameworks Influencing Obesity

3

1.1 ISSUE HISTORY

Overweight and obesity has been a national concern for over two decades. Data collected

by the Canadian Community Health Survey (CCHS) in 2014 revealed that 54% of

Canadian adults are defined as overweight or obese.1 In the same survey, 68% of adults in

Newfoundland and Labrador (NL) were classified as overweight or obese, this was the

highest rate seen in all of Canada’s provinces and territories.1

Overweight and obesity is associated with numerous chronic health conditions, including

but not exclusive to, type 2 diabetes, cardiovascular disease, hypertension, osteoarthritis

and certain types of cancer.2 According to the CCHS, 25% of Newfoundland adults suffer

from hypertension, and 9% have been diagnosed with diabetes; not only do these rates

exceed the national average, they are the highest provincial/territorial rates in Canada.1

Further, 40% of Newfoundland’s population live in rural or remote communities.3 It has

been consistently documented that people living in rural and remote communities have

poorer health status than Canadians who live in urban areas.4 Canadians living in rural

areas are more likely to be of poor socio-economic status, practice less healthy behaviors

and to have higher overall mortality rates than urban residents4, making them even more

vulnerable to the negative health outcomes related to overweight and obesity.

Overweight and obesity is a complex issue with multiple contributing factors. In the

simplest description, an individual will become overweight or obese when their energy

intake consistently exceeds their energy output.5 Numerous factors contribute to

4

overweight and obesity but one of the most influential contributors is poor nutrition. It

has been widely accepted that poor diet quality, including diets that are high in fat, sugar,

salt and contain fewer nutrient dense foods such as fruits, vegetables and whole grains,

can lead to the energy imbalance that causes overweight and obesity.5 Currently,

Newfoundland has the second lowest fruit and vegetable consumption in Canada; only

26% of individuals 12 and older consuming at least 5 servings of vegetables and fruit per

day.1

1.2 THEORETICAL FRAMEWORKS OF FACTORS INFLUENCING OBESITY

What public health researchers have become increasingly interested in is which factors

contribute to poor dietary behaviors.6 Peripheral to the biological reasons people eat (i.e.

hunger, appetite and taste), various other factors of the social and built environments can

independently and collectively work to influence individual food choice.6 Researchers

across disciplines have dedicated efforts to understanding how these factors work and

interact as determinants of dietary behavior.

Using an ecological framework, Story et.al, illustrated the multifactorial influences on

dietary behavior as a nested model (Figure1-1).6 They have defined the most exterior

determinants as macro-level environments and sectors which include societal and cultural

norms and values as well as legislative, regulatory or policy action within government

and industry. Within macro-level environments are physical environments (home, work,

school) and community environments and the access, availability, barriers and

5

opportunities to interact with food within these settings. Within physical environments

are social environments, an individuals’ family, friends and peer networks, social norms

and social supports. Finally, within social networks there are the individual level or

personal factors, including biological and socioeconomic demographics, attitudes,

preferences, and knowledge and skills surrounding food.6

Figure 1-1 Ecological framework of multifactorial influences on diet

Story, M. et al. 2008. Reproduced with permission.

6

Recognizing how these factors and systems have changed to impact the food environment

is imperative to the investigation into individual food choice and understanding the

obesity epidemic. Macro-level systems have evolved and have become more

sophisticated; cultivating an increase in the marketing, availability and affordability of

highly processed, high fat and high sugar food and drink, larger portion sizes and

increased dependence on and accessibility to low nutrition fast foods and ready-made

items.6 As a result, these changes are interacting and influencing physical environments,

social systems and individual level preferences and behaviors.6

Lytle conceptualized this relationship between environmental factors and individual and

social factors using a visual schematic (Figure 1-2).7 The model demonstrates that as the

availability and accessibility of healthy options become more restricted, the stronger the

influence the physical environment may have on individual food choice. Alternatively, as

availability and accessibility increase, the influence of the physical environment may

decrease and the influence of individual behavior and social factors may increase.7

7

Figure 1-2 The relationship among individual, environmental and social factors

Lytle, L. 2009. Reproduced with permission.

Acknowledging the significant transformation in the physical environments and the

concept that in order to observe change in individual behaviors there must be supportive

environments in which to make healthy choices, Glanz and colleagues 8 took a narrower

scope, inspecting the elements of the physical food environment while also considering

how it is affected by the multiple levels of interacting influences. They divided the

physical food environment into 3 components: the community nutrition environment,

defined as the type and location of food outlets; the organizational nutrition environment

including those within work, schools and home; and the consumer nutrition environment,

involving the availability of foods and the price, promotion and placement within an

outlet (Figure 1-3).8

8

Figure 1-3 Model of community nutrition environment

Glanz, K. et, al. 2008. Reproduced with permission.

This research has been conducted under the lens of the Glanz model. Combining the

scope of the consumer and community nutrition environment, we have investigated the

retail (consumer) food environment within the context of the geographic accessibility and

the in-store availability, quality and price of nutritious food across a defined area or

community.

9

CHAPTER 2: Existing Evidence & Local Context

10

2.1 EXISTING EVIDENCE

In response to the concerning rates of obesity and other diet related diseases, the study of

the physical food environment has commanded attention among nutrition and public

health researchers in recent years.9 A 2013 review by Health Canada9 included as many

as 81 food environment articles, all which have been published since 2000. A more recent

review, published in 201610, included 88 Canadian food environment articles, 75% of

which had been published in the last 5 years. Unfortunately, as recognized in the Health

Canada review and by others, the majority of literature available on food environments

has been conducted within low income urban areas and few researchers have considered

the food environment in a rural context.9,10

The taxonomy of rurality is multifaceted and this complexity will be explored at more

detail throughout this thesis. However, at the broadest interpretation, and for the purpose

of this thesis, rural areas are anything outside of what is generally considered urban. Rural

communities differ from their urban areas because they have smaller populations over a

larger area of land.4 Rural residents are further from the metropolitan area and often have

to travel a significant distance for some of their basic necessities.4 Public transportation is

often limited in rural communities4, therefore, residents that do not have access to a

vehicle may have increased reliance on nearby food retailers. As a consequence of these

disparities, it may be inappropriate to generalize food environment research conducted in

urban areas to rural food environments.

11

The research that has been done within rural food environments has indicated geographic

access to supermarkets and grocery stores is significantly limited in rural communities

and there is a higher density of small convenience stores compared to their urban

counterparts.11 This is concerning because of evidence indicating that most often, larger

supermarkets and grocery stores have a larger variety of affordable nutrient dense foods

while smaller convenience and corner stores have a higher availability of less nutritious,

energy dense, snack foods.11 When healthier items are available in small stores, they tend

to be of a higher price and lower quality.11

Research conducted by Leise and colleagues surveying store density and store

distribution in a primarily rural South Carolina county found that 60% of rural

communities in the county had no supermarket and 40% had no grocery store, while 49%

of the county’s convenience stores were located in the rural areas.12 Comparably, Vilaro

et al. measured the food environment in rural Florida and found that 72% of the food

retailers in the community of interest were convenience stores.13 Dean et al. analyzed the

community food environment by distance and documented that it was on average 8.9

miles to the nearest supermarket or supercenter in rural areas and only 3.0 miles in urban

areas.14 Similarly, Michimi and Wimberly found it was on average 8.1 miles to the nearest

supermarket or supercenter in rural areas compared to 3.6 miles in urban areas 15.

This geographic disparity is troubling because research has shown the lack of access to

supermarkets and grocery stores can have a significant impact on the types of food

available to those in rural communities.11 The work by Leise et al. also examined food

12

availability in rural supermarkets, grocery stores and convenience stores, revealing that

the availability of healthy foods including low-fat/fat-free milk, apples, high fibre bread

and eggs was considerably higher in supermarkets and grocery stores.12 Only 2% of

convenience stores offered any produce, only 2% offered low-fat/non-fat milk and only

4% offered high fiber bread.12 The most frequently offered healthy items included low

fibre bread (86%) and whole milk (68%) 12 A similar study by Jithitikulchai et al. had

comparable findings.16 All of the surveyed items were available at all urban supermarkets

but fewer were available in rural supermarkets.16 Canned vegetables were the most

common type of vegetables sold in rural convenience stores, only 3% had fresh apples,

bananas, onions and tomatoes and even fewer had oranges and carrots.16 Very few of the

rural convenience stores sold fresh meat or poultry and skim milk and whole wheat

options were rare.16 Both studies also measured the price of healthier options and found

that the average cost of healthy items (when available) including fresh produce, eggs,

low-fat/fat free milk and whole grains, were substantially higher at convenience stores

than at supermarkets and grocery stores.12,16

Using a comprehensive approach, Vilaro et al. examined the availability, price and

quality within retail stores using scoring system.13 Supermarkets scored on average 21.33

out of a possible 30 points for the availability of healthy items, while convenience stores

had a mean score of 2.88.13 When comparing price between healthy and unhealthy items,

supermarkets scored an average of 4.67 out of 17 possible points while convenience

stores scored 0. When measuring quality of fresh produce, supermarkets scored an

13

average 6.00 out of 6 possible points compared to an average of 3 points in convenience

stores.13

The evidence provided in this sample of food environment literature demonstrates the

significant barriers to food access in rural communities. The geographic access to

supermarkets in rural communities is limited and as a result rural residents may be forced

to rely on smaller convenience stores for their groceries. Convenience stores were shown

to have limited availability of healthy items and what was available was considerably

more expensive than they were at larger supermarkets and grocery stores.

2.2 THE LOCAL CONTEXT

These findings are of particular concern to Newfoundland due to the significant

proportion of the population residing in rural or remote areas.3 Research has also shown

Newfoundland has the highest number of convenience stores per capita of any

province/territory in Canada (~1 per 495) and the highest proportion of convenience

stores in rural areas (72%), suggesting convenience stores play a larger role in food

choices in Newfoundland than elsewhere in Canada.17 Unfortunately, despite the evidence

that poor food environments exist in Newfoundland and Labrador, few provincial policies

and strategies are in place to improve access to healthy nutritious food, demonstrating an

urgency for action.18

14

Researchers at the Food Policy Lab at Memorial University acknowledged the need for

investigation and have launched the Healthy Corner Stores NL project.17,19 The aim of the

project is to implement a pilot program to define and test the potential for healthier foods

in Newfoundland convenience stores.17,19 To inform their project, researchers have

conducted Nutrition Environment Measures Surveys (NEMS), collecting data on the

availability, quality and price of healthy foods in 78 convenience stores across the Avalon

Peninsula.19 To date, only preliminary analysis has been conducted on the availability of

select items and further analysis is needed to get comprehensive perspective of all

relevant features of the food environment.19

15

CHAPTER 3: Evidence to Action: Measuring the Food Environment and an Introduction

to the Nutrition Environment Measures Survey

16

3.1 MEASURING THE FOOD ENVIRONMENT

In order to encourage action and influence change, it is most effective to produce

evidence that is meaningful to policymakers, retail owners and other key stakeholders,

guiding them in prioritized and informed decision making. This concept is referred to as

the element of “response” in a 3 tiered model of the factors to consider when deciding on

a food environment assessment strategy proposed by Dr. Leia Minaker, a food

environment researcher at the University of Waterloo.20 In the model she refers to three

considerations that should drive food environment assessment: resources, relevance and

response. Resources refers to the consideration of what is feasible and appropriate within

the community of interest before conducting a food environment assessment. Relevance

refers to how food environment problems are related to an outcome of interest. Finally,

response refers to the ability of policy-makers to find meaning out of and act on the

evidence presented.20

Despite the increasing interest in the food environment, there are limited recognized and

standardized methods available to measure the key constructs of the food environment in

a way that will portray the overall healthfulness of a retailer and identify priority areas

where change is needed.21 A review by Kelly et.al in 2009, explored the available

methods and measures of the food environments and identified an extensive list of

classification tools, instruments and indices that are being used to measure and define

community and consumer food environments.21 The Health Canada review “Measuring

the Food Environment in Canada” has also highlighted the number of existing food

17

environment assessment methods, referring to a database of over 500 assessment tools

compiled by the United States National Cancer Institute.22 However, as acknowledged by

Dr. Minaker, “There are thousands of food environment features that are measurable, but

not all measures are equal in terms of their ability to raise awareness or inform policy

priorities.”.20

Methods relevant to the retail food environment can be classified into two categories:

consumer (retail) nutrition environment measures and community nutrition environment

measures, corresponding with the priority concepts of the food environment as modeled

by Glanz et al 10,21 Community nutrition environment measures are defined as those

examining geographic access by measuring the density, proximity or variety of food

retailers within a defined area. 9,10, Based on the evidence that grocery stores and

supermarkets have a higher availability of nutritious foods compared to convenience

stores, some researchers define the healthfulness of food environments using exclusively

geographic access. Many researchers will use a pre-existing database of food retailers

developed by related organizations, such as the provincial food establishment license

database.10,21 Often these databases will code stores by store type, one of the most

commonly used systems is the North American Industry Classification System (NAICS).

Once the geographic coordinates of the relevant food retailers have been determined, the

distance, density or variety of store types within the study area can be measured.10,21 The

disadvantage of this method is that it relies on the accuracy of existing databases and the

assumption that findings in previous literature are generalizable.21 The appeal to this

method is that it is simpler and requires minimal resource and time, allowing researchers

18

to measure the food environment in large areas more efficiently. This approach is most

effective for environmental scans or needs assessments; it allows researchers to identify

whether food environment disparities exist, but is limited in its ability to measure or

describe the key constructs of the food environment that are necessary to guide

community level action.

Direct observation of the consumer food environment has been considered a more

sophisticated and robust measure of the food environment.21,23 Consumer food

environment measures involve assessment of the availability, quality and price of food

within stores.10,21 A variety of different tools are available to measure the constructs of

the consumer food environment. Market baskets have been used to measure food

affordability, checklists and surveys to collect data on the availability, price or quality of

foods and shelf base measures which assess the prominence of food within stores.10,21

These approaches are more resource intensive, however, the benefit of working in rural

areas that there are few and smaller retailers, making it more feasible to measure the

actual availability, price and quality of foods within the community.20

3.2 NUTRITION ENVIRONMENT MEASURES SURVEY

Among the most widely used consumer nutrition environment measurement tools is the

Nutrition Environment Measurement Survey (NEMS). NEMS is an observational survey

designed by Glanz et al., to survey the consumer nutrition environments including the

availability, pricing and quality of food within retail food environments.24 The items

19

included in the survey were selected based on the types of food that contribute the most

fat and calories to an average American diet and those that are most recommended for

healthful eating. The development of the food list was informed by publications from

various relevant federal government agencies, health professional organizations and

researchers.24 Items in the NEMS are organized into 11 categories and with the exception

of fruit and vegetables, each category has “regular” options and “healthful” options.

Points are awarded for having the healthy options of the audited items and for some items

additional points are awarded for having increasing variety.24 Price is measured

comparatively, therefore, points are given for having the healthy option priced lower than

the regular option for each audited item and points are subtracted if the healthful option is

priced higher than the regular option.24 Quality is measured for fresh fruit and vegetables

only and points are awarded based on the proportion of all fruit items and the proportion

of all vegetable items that have been recorded as acceptable.24 Scores are calculated for

availability, quality, and price, and an overall score combines the 3 dimensions. A higher

score indicates higher quality, availability, and/or lower prices for the healthier items.24

Although there is no gold standard in food environment measures, NEMS frequently

serves as the “proxy” gold standard in food environment evaluation tools in North

America.25,26,27,28 The NEMS survey was tested in 88 stores and found high values for

both interrater and test-retest reliability.24 As obesity and diet related disease remain an

urgent public health crisis, food environment research continues to grow, propagating

research using the NEMS methods. A non-exhaustive list of publications on the NEMS

website documents over 60 journal articles using the NEMS method since 2008.29 Since

20

its development in 2007, the observational Nutrition Environment Measures Survey has

been adapted for use in restaurants, grocery stores and supermarkets, corner stores, and

vending machines.24,29 Most relevant to the retail food environment are the NEMS-S

(Nutrition Environment Measures for Stores), which is designed to evaluate grocery

stores, supermarkets and corner stores and NEMS-CS (Nutrition Environment Measures

for Corner Stores), a modified version of NEMS-S designed to evaluate corner stores and

convenience stores.24,29 With the exception of two added categories, NEMS-S and NEMS-

CS are identical, therefore are frequently referred interchangeably as NEMS-S.24,29

One of the main benefits of a NEMS-S score is that it can provide both sub scores for

availability quality and price and composite scores. Although more resource intensive,

this approach measures the key constructs of the food environment and allows for

comprehensive comparison among and within store type and is valuable for research

investigating analytical correlations.

Unfortunately, the standard methods used to analyze data collected using NEMS-S are

less accessible in a community setting; without context, the NEMS-S scores may have

little meaning to decision makers, retail owners or their consumers.20 However, with the

expansion of food environment research, many users have begun to adapt the NEMS-S by

modifying the items audited, altering the scoring rubric, or tailoring the method of

analysis to suit their study objective and less is known about the efficacy of these

approaches in producing evidence that is more suitable for dissemination to decision

makers and stakeholders.

21

Researchers have modified the NEMS-S survey to reflect culturally appropriate diets or

locally available foods, specific demographics, certain store programs or policies and

specific diets or food groups. Some users make minor changes: Lasley et al. modified

their survey to reflect local produce availability in rural Iowa by changing the type of

pears and tomatoes included in the survey30, while others make significant changes.

Martins et al. made extensive changes to the original survey to create a NEMS-S

appropriate for urban Brazil.31 Changes involved creating three main groups: unprocessed

or minimally processed foods (healthier), refined ingredients for use in culinary

preparations and the food industry (intermediate), and ultra-processed products (less

healthy).31 Paek et al. adapted their survey to evaluate the changes made in stores after the

implementation of healthy corner store intervention (FIT) in Detroit Michigan32 and

Andreyeva, et al. modified the survey to focus on WIC (nutrition assistance program)

approved foods33. Some researchers choose to look at only certain food groups or

nutrients, Johnson et al. was interested in fruit and vegetable consumption therefore only

collected NEMS-S data on fresh produce34, while Hermstad et al. modified the survey to

only collect data on items that were low fat or lean.35

NEMS-S users are not only adapting the items included in the survey, but they are also

making changes in how it is analyzed and evaluated. Some users have decided not to use

the score, analyzing the data using descriptive statistics only. Andreyava, et al. reported

their data by the percentage of stores within the community that had each item audited,

reported mean price of each item and calculated a market basket price and quality was

reported as the percent of stores that had overall fair, good, or excellent quality produce.33

22

Others have collected data using the NEMS-S survey but adapted the scoring system.

Franco et al. used a Healthy Food Availability Index (HFAI).36 HFAI considers only the

availability of healthy foods, ranging from 0-27 with higher scores indicate greater

availability of healthy foods. They then analyzed the scores by categorizing them into

tertiles: high, medium and low based on the sample scores and reported the percentage of

stores in each community falling within each category.36

While research reviewing food environment measures has been done in the past, none to

our knowledge have focused exclusively on the NEMS-S method and less is known about

how users may be adapting the Nutrition Environment Measures Survey to evaluate the

food environment. Since NEMS-S is currently the serving as the gold standard in

consumer food environment research and publications featuring the NEMS-S method are

increasing, it is imperative that we understand the ways in which the survey is being used,

modified, and analyzed to adopt key learnings and identify methods of analysis that are

more meaningful in a community context and will expose priority areas where change is

needed.

23

CHAPTER 4: Research Rationale, Objectives & Design

24

4.1 RATIONALE

NL currently has the second lowest fruit and vegetable consumption in Canada and has

the highest prevalence of some of the most leading diet related diseases, including

obesity, diabetes and hypertension.1 Recently, increased attention has been given to the

role of the food environment on diet quality and the development of diet related diseases;

suggesting, those who live in disadvantage food environments where the accessibility,

availability and affordability of nutritious food is limited, may be more vulnerable to the

development of diet related diseases.10

Community food environment research has shown that geographic access to supermarkets

and grocery stores is significantly limited in rural communities and there is a higher

density of small convenience stores compared to their urban counterparts.11 This has been

proven to be true in Newfoundland, where 72% of rural food retailers are convenience

stores.18 Food environment measures have also indicated that convenience stores in rural

communities are less likely to stock quality, affordable, nutritious food.11 These findings

are troubling considering that 41% of the provincial population living in rural or remote

communities3 where there may little to no access to nutritious foods. Unfortunately, there

are currently no direct provincial programs or policies in place aimed at improving

equitable access to healthy foods in rural Newfoundland.18

Recognizing the need for further investigation into the food environment in

Newfoundland and Labrador researchers from the Food Policy Lab at Memorial

25

University launched the Healthy Corner Store Newfoundland project (HCSNL).19As part

of the project, NEMS-S were conducted in 78 stores across the Avalon Peninsula. To

date, only preliminary analysis has been conducted on the data and more work is needed

to describe and define the healthfulness of the retail food environment in

Newfoundland.19

Although standard approaches in analyzing and evaluating NEMS-S research serve a

valuable purpose in the assessment of the food environments, they are less meaningful in

a community context.20 In order to support action, it is necessary to produce evidence that

is effective in selecting an appropriate process through which to make change and will

work to inform prioritized strategies and interventions. Over 60 articles using the NEMS-

S method have been published since 2008 and to our knowledge there have been no

synthesis of the ways in which the survey has been used, modified, and analyzed.29

For these reasons, the objectives of this research were to review and synthesize the

methods used in the analysis and evaluation food environment data collected using the

NEMS-S method in rural communities. By applying 2 models gathered through the

scoping review (healthy food assessment only, healthy and unhealthy food assessment) to

the secondary NEMS dataset collected by HCSNL, our goal was to compare the outcomes

produced by each model. While working toward achieving this goal, we simultaneously

evaluated the food environment across the Avalon Peninsula. Finally, we proposed and

analyzed two fiscal based policy options to improve availability and affordability of

nutritious food in rural Newfoundland and Labrador.

26

The ultimate purpose of our project was to identify approaches to describe rural food

environments in a format that is transferable to prioritized policy development and will

lead to the implementation of impactful change and improvement within rural

communities, while providing necessary and accessible evidence needed to guide

informed decision making regarding strategies and interventions aimed at improving the

availability and accessibility of healthy nutritious food in Newfoundland and Labrador.

4.2 OBJECTIVES

1. Identify the factors and methods being used to describe and define healthfulness of the

rural consumer food environment

2. Determine the significance and implications of assessing the consumer food

environment using healthy food only (and not unhealthy)

3. Identify and analyze potential policy options to increase the availability and

accessibility of nutritious foods in rural retail food outlets in Newfoundland and Labrador

4.3 DESIGN

To achieve our research objectives, we have conducted a scoping review of the literature

evaluating the healthfulness of retail food environments in rural communities using the

Nutrition Environment Measures Survey. The literature was reviewed and synthesized to

27

identify key themes and methodological approaches. Using the data collected through the

review, we identified 2 models for the analysis and evaluation of objective consumer food

environment measures in rural communities. Next, we conducted secondary analysis of

cross-sectional quantitative data previously collected as part of Healthy Corner Store NL

project by applying each model of analysis to the dataset and describing and comparing

the outcomes. Finally, using individual reflection based on the knowledge gained in the

previous steps, supplemented by a critical examination of relevant scientific and grey

literature, we proposed and analyzed two potential policy options to improve access to

nutritious foods in rural Newfoundland using the framework developed by National

Centre for Healthy Public Policy.

4.4 ETHICS

The scoping review and policy analysis involve no human participants, personal or

sensitive information, therefore employs minimal ethical risks. Analysis of secondary

data collected as part of the Healthy Corner Stores NL has been previously granted ethical

approval by the Newfoundland and Labrador Health Research Ethics Board (HREB

#15.145).

28

REFERENCES

1. Statistics Canada. Canadian community healthy survey-annual component summary tables [Internet]. Statistics Canada. 2015. Available from: http://www5.statcan.gc.ca/COR-COR/COR-COR/objList?lang=eng&srcObjType=SDDS&srcObjId=3226&tgtObjType=CST

2. Public Health Agency of Canada, Canadian Institute of Health Information. Obesity in Canada [Internet]. Ottawa Ontario: Public Health Agency of Canada and the Canadian Institute for Health Information; 2011 [cited 2017 Apr 26]. Available from: https://secure.cihi.ca/free_products/Obesity_in_canada_2011_en.pdf

3. Statistics Canada. Population, urban and rural by province and territory (Newfoundland and Labrador) [Internet]. 2012. Available from: https://www12.statcan.gc.ca/census-recensement/2011/ref/dict/geo042-eng.cfm

4. Crosby R, Wendel M, Vanderpool R, Casey B. Rural populations and health determinants, disparities, and solutions. 1st ed. Wiley; 2012.

5. Wright S, Aronne L. Causes of obesity. Abdom Imaging. 2012 Oct;37(5):730–2.

6. Story M, Kaphingst K, Robinson-O’Brien R, Glanz K. Creating healthy food and eating environments: Policy and environmental approaches. Annu Rev Public Health. 2008;29:253–72.

7. Lytle L. Measuring the food environment: state of the science. Am J Prev Med. 2009 Apr;36(4):S134–44.

8. Glanz K, Sallis J, Saelens B, Frank L. Healthy nutrition environments: Concepts and measures. Am J Health Promot. 2005 Jun;19(5):330–3.

9. Health Canada. Measuring the food environment in Canada [Internet]. Ottawa Ontario; 2013 Oct. Available from: http://publications.gc.ca/collections/collection_2013/sc-hc/H164-155-2012-eng.pdf

10. Minaker L, Shuh A, Olstad D, Engler-Stringer R, Black J, Mah C. Retail food environments in Canada: A scoping review. Can J Public Health. 2016 Jun;107(1):5344.

11. Larson N, Story M, Nelson M. Neighborhood environments: Disparities in acess to healthy foods in the U.S. Am J Prev Med. 2008 Jan;36(1).

29

12. Liese A, Weis K, Pluto D, Smith E, Lawson A. Food store types, availability and costs of food in a rural environment. J Am Diet Assoc. 2007 Nov;107(11):1916–23.

13. Vilaro M, Barnett T. The rural food environment: A survey of food price, availability, and quality in a rural Florida community. Food and Public Health. 2013;3(3):111–8.

14. Dean W, Sharkey J. Rural and urban differences in the associations between characteristics of the community food environment and fruit and vegetable intake. J Nutr Educ Behav. 2011 Nov;43(6):426–33.

15. Michimi A, Wimberly M. Associations of supermarket accessibility with obesity and fruit and vegetable consumption in the conterminous United States. Int J Health Geogr. 2010 Oct;9(49).

16. Jithitikulchai T, Dean W, Sharkey J. Variations in the availability and price of healthier food options by store type and urban-rural setting. 2012 Dec;7(4):381–400.

17. Memorial University. Healthy corner stores NL [Internet]. 2015. Available from: http://www.yaffle.ca/a/viewProject/id/1936

18. Newfoundland and Labrador. Department of Health and Community Services Strategic Plan 2014-2017 [Internet]. n.d. Available from: http://www.health.gov.nl.ca/health/publications/Department_of_Health_and_Community _Services_Strategic_Plan_2014-17.pdf

19. Mah C. Healthy Corner Stores NL Milestone Report.

20. Minaker L. Retail food environments in Canada: Maximizing the impact of research, policy and practice. Can J Public Health. 2016 Jun;107(1):5632.

21. Kelly B, Flood V, Yeatman H. Measuring local food environments: An overview of available methods and measures. Health Place. 2011 Nov;17(6):1284–93.

22. National Cancer Institute Division of Cancer Control and Populations Sciences. Measures of the Food Environment. [Internet]. Available from: http://epi.grants.cancer.gov/mfe

23. Cobb L, Appel L, Franco M, Jones-Smith J, Nur A, Anderson C. The relationship of the local food environment with obesity: A systematic review of methods. study quality and results. Obesity. 2015 Jul;23(7):1331–44.

24. Glanz K, Sallis J., Saelens BE, Frank L. Nutrition Environment Measures Survey in stores (NEMS-S): Development and evaluation. Am J Prev Med. 2007 Apr;34(4):282–9.

30

25. Minaker L, Raine K, Wild T, Nykiforuk C, Thompson M, Frank L. Construct

validation of 4 food-environment assessment methods: Adapting a multitrait-multimethod matrix approach for environmental measures. Am J Epidemiol. 2014 Feb;179(4):519–28.

26. Glanz K, Johnson L, Yaroch A, Phillips M, Ayala G, Davis E. Measures of retail food store environments and sales: Reviews and implications for healthy eating initiatives. 2016 Apr;48(4):280–8.

27. Moore L, Diez Roux A, Franco M. Measuring availability of healthy foods agreement between directly measured and self-reported data. Am J Epidemiol. 2012 May;175(10):1037–44.

28. Minaker L, Raine K, Wild T, Nykiforuk C, Thompson M, Frank L. Objective food environments and health outcomes. Am J Prev Med. 2013 Sep;45(3):289–96.

29. Nutrition Envionrment Measures Survey [Internet]. [cited 2017 Apr 27]. Available from: http://www.med.upenn.edu/nems/

30. Lasley E, Litchfield R. Fresh produce in rural Iowa: Availability and Accessibility. J Hunger Environ Nutr. 2008 Oct;2(2–3):5–17.

31. Martins P, Cremm E, Letite F, Maron L, Scagliusi F, Oliveria M. Validation of the Nutrition Environment Measurement Tool for Stores (NEMS-S) in an urban area of Brazil. J Nutr Educ Behav. 2013 Dec;45(6):785–92.

32. Paek J Hye-Jin, Oh J Hyun, Jung J Yumi, Thompson J Tracy, Alaimo J Katherine, Risley J John, et al. Assessment of a Healthy Corner Store Program (FIT Store) in Low- Income, Urban, and Ethnically Diverse Neighborhoods in Michigan. Fam Community Health. 2014;37(1):86–99.

33. Andreyeva T, Luedicke J, Middleton A, Long M, Schwartz M. Positive influence of the revised special supplemental nutrition program for women, infants, and children food packages on access to healthy foods. J Acad Nutr Diet. 2012 Jun;112(6):850–8.

34. Johnson J, Nobmann E, Asay E. Factors related to fruit, vegetable and traditional food consumption which may affect health among Alaska Native People in Western Alaska. Int J Circumpolar Health. 2012 Mar;71(0):183–8.

35. Hermstad A, Swan D, Kegler M, Barnette J, Glanz K. Individual and environmental correlates of dietary fat intake in rural communities: A structural equation model analysis. Soc Sci Med. 2010 Jul;71(1):93–101.

31

36. Franco M, Diez Roux A, Glass T, Caballero B, Brancati F. Neighborhood characteristics and availability of healthy foods in Baltimore. 2008 Dec;35(6):561–7.

SECTION II

Manuscript Series

32

CHAPTER 5: The Nutrition Environment Measures Survey in Rural Communities: A

Scoping Review

33

AUTHORSHIP STATEMENT

TITLE: The Nutrition Environment Measures Survey in rural communities: A scoping

review

AUTHORS: 1. Rebecca Harris (Lead Author), MSc Student, Community Health and Humanities, Memorial University of Newfoundland and Labrador 2. Dr. Leia Minaker, Assistant Professor, School of Planning, University of Waterloo 3. Dr. Catherine Mah, Associate Professor, School of Health Administration, Dalhousie University

PUBLICATIONS: A version of this manuscript has been submitted for publication in the

Journal of Hunger and Environmental Nutrition.

Contributions

RH and CM devised the idea for the review and conceptualized the design and

implementation plan. RH conducted literature review. RH created the search strings which

were reviewed by CM. RH conducted the database search and screened articles for

inclusion, CM conducted an abstract review of the final list of articles to confirm inclusion.

RH developed coding template for data extraction, CM reviewed. RH conducted full text

review and coded the articles. CM conducted full text review to confirm coding. RH

conducted data synthesis and interpretation, CM reviewed. LM was a key consultant

throughout the literature search, data extraction, synthesis and interpretation, contributing

thoughts and feedback.

34

All drafts of the manuscript (including all sections) were written by RH with initial,

intermediate, and final edits and suggested revisions by CM. Manuscript was sent to LM

for intermediate and final editing and suggested revisions. RH devised final draft and

completed citation and works cited. RH formatted and prepared the manuscript for

submission for publication.

35

5.1 ABSTRACT

The Nutrition Environment Measures Survey (NEMS-S) is one of the most widely used

food environment assessment tools. This systematic literature review focuses on

evaluation of rural retail food environments using the NEMS-S, with particular attention

to methodological adaptations. The review demonstrates the use of heterogeneous

terminology combined with a lack of transparency in some of the language used to

describe key constructs of the research methodology including: community classification,

store enumeration and store classification. This is important for rural food environments

which appear to vary more widely than urban settings. To increase comparability across

the literature, future research should work to increase the transparency in the

methodology; providing explicit definitions and rationales.

36

5.2 INTRODUCTION

Research on the retail food environment encompasses the community nutrition

environment (the type and location of food outlets within a defined geographic area)1, and

the consumer food environment (the availability and marketing of foods and beverages

encountered in food outlets).1 The retail food environment may contribute to population

dietary quality and health outcomes.2,3 Direct observation of the consumer food

environment is considered among the most robust measures of the food environment.4-9

The Nutrition Environment Measures Survey for Stores (NEMS-S)10 currently serves as

the gold standard in consumer food assessment tools.9,11,12 NEMS-S is a checklist measure

designed by Glanz and colleagues for store audits of availability (including variety), price

and quality of food and beverage products.10

The NEMS-S was designed to evaluate grocery stores and supermarkets.10 The NEMS-

CS (Nutrition Environment Measures for Corner Stores) is based on the NEMS-S,

designed for corner stores and convenience stores, and contains two additional

measures.13 For the purposes of this article, and because they are closely related, we will

refer to both collectively as NEMS-S. The NEMS-S collects data on both regular and

healthier (e.g., lower fat, lower sugar) alternatives for foods and beverages that contribute

the most fat and calories to the American diet.10,14

The NEMS-S scoring system consists of three sub-scores reflecting food availability,

price and quality, summed in a composite score.10,14 Points for availability are awarded

37

for healthful alternatives to regular items, and for some items, additional points are

awarded for variety.10,14 Price is measured comparatively, assessing whether the healthful

option is priced equal to or lower than a given regular item.10,14 Quality is measured for

fresh produce only, based on the proportion of items that have been recorded as

acceptable (i.e., not withered, bruised, or wilted).10,14

An important area of consumer food environment research is the emerging work in rural

areas, where vulnerability to diet related disease is greater.15,16 Evidence suggests that the

variability in the quality of food environments in rural and remote regions is greater than

in urban areas15 and NEMS-S research is a valuable way to contextualize these

differences at the consumer interface inside stores.

One of the challenges of existing NEMS-S research is that the scores may not have

practical significance for decision makers, retail store owners, or their consumers.17 In

order to plan interventions and influence healthy changes in the food environment, it is

important to understand how evidence can be made accessible to knowledge users for

decision making. However, with the expansion of the food environment research

literature, many users have begun to adapt the NEMS-S by modifying the items audited,

altering the scoring rubric, or tailoring the method of analysis to suit their study

objective14,18 and it is unclear how these adaptations capture the variables necessary to

guide intervention planning and policy development.

38

For these reasons, the objective of this research was to systematically review the literature

to synthesize how features of the NEMS-S are used to measure the rural consumer food

environment. Our secondary objective is to contribute to efforts to refine methods to

analyze and evaluate rural and remote food environment data in formats that are

accessible to knowledge users.

5.3 METHODS

This review follows the framework of a scoping review as defined by Arksey and

O’Malley.19 We began first by posing our research question and objectives, next we

identified relevant literature, selected the applicable studies, and then extracted the data.

Finally, we summarized and collated the data into key themes.

Data Sources

Literature was collected using database search, reference search and expert consultation

in September of 2016. The database search was conducted in PubMed, Web of Science,

and Scopus using predetermined search strings (Figure 5-1).

Study Selection

Peer-reviewed journal articles published from 2007-2016 were eligible for inclusion to

correspond with articles published after the initial publication of the NEMS-S method.10

All types of study designs gathering and analyzing data on the rural consumer food

environment using NEMS-S/NEMS-CS were included. For the purposes of this review,

39

any study that defined the assessed community as “rural” or “remote” was included. Grey

literature, including reports, editorials, and news articles, were excluded. Research

exclusively assessing restaurants, hospitals, schools, or other primarily organizational

food environments, were excluded.

A total of 55 papers were screened in through the database searches. First, 18 duplicates

were removed. Next, five were identified as review articles and were scanned for relevant

references and removed from the collection. Finally, 2 papers were added based on expert

consultation, resulting in 34 articles for full text review. During full text review, 16

records were excluded because they did not use the NEMS-S. One record was removed

because it did not address rural areas (Figure 5-1).

Data Extraction

A total of 17 articles met the criteria for inclusion in the review. Two researchers

independently reviewed each paper for key concepts and themes to develop a common

code template, which was used to independently code the literature, then all team

members deliberated upon the coding until consensus was reached.

Data was extracted on study design, jurisdiction, store type, store enumeration methods

(the process used to identify and locate the audited food outlets), survey adaptions,

reported NEMS-S features, and scoring methods. All included studies were of the same

design: cross-sectional, descriptive analyses of the consumer food environment. Where

40

possible, studies were also coded as evaluation or formative research, including baseline

and outcome studies associated with evaluating a program or policy.

Figure 5-1 Review Literature Selection

!(Rural!OR!remote)!AND!(Food!environment)!AND!(Nutrition!

Environment!Measures!Survey!OR!NEMS!OR!Asses*!OR!evaluat*)!AND!(availab*!OR!access*!OR!afford*)!

!Articles!included!in!analysis!after!full!text!review!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

n=17!

!Excluded!!Not!NEMSDS!

n=16!

Excluded!!Not!Rural!

n=1!

PubMed!n=199!

Scopus!!n=614!

Web!of!Science!n=!183!

Articles!after!title!and!abstract!screen!(duplicates!and!review!articles!

removed)!!n=32!

PubMed=10!Scoups=8!Web!of!Science=17!!

Iden

tifica

tion+

Screen

ing+

Eligibility+

Inclu

ded+

Articles!for!full!text!review!!!n=34!

!

Recommended!!by!expert!!consult!n=!2!

41

5.4 RESULTS

Study Design

All studies were cross-sectional in nature. Three studies assessed food environment

associations with area-level diet or health factors20-22. Fourteen articles23-36 were

descriptive research, describing food environment features by store type (n=8)24-

26,28,30,31,34,36, community size or rurality (n=7)23-25, 29,30,33,34, or community or population

sociodemographics (n=2)25,28. Of these, eight were evaluations32, 34-36 or formative

research27-29,31.

Jurisdiction

All 17 studies were conducted in the United States (US). Eleven articles exclusively

observed rural areas20-22,26-28,30,31,33,35,36 while the remaining six articles examined both

urban and rural areas23-25,29,32,34. Only nine articles provided an explicit explanation for

how they defined rural 21,23,24,27,30,32,33,34,36 (Figure 5-1). Of these, three studies used the

US Department of Agriculture rural-urban continuum codes21,23,33. Two articles32,34 used

the local school National Center for Education Statistics rural/urban designation to

classify the community. One article36 used a local health department definition for rural

designation, and two used the federal Census definition for rural24,30.

Store Type

Grocery stores/supermarkets were the most frequently (n=13)20-26, 28-31, 33, 35 surveyed

store type, followed by convenience/corner stores (n=9)20-22, 24, 26, 28-31. Other store types

42

examined included dollar stores, pharmacies, supercentres, and specialty

stores21,22,24,25,26,28. Store type was identified differently across the reviewed literature:

three studies defined store type based on revenue30,32,34; one classified stores based on

size, number of cash registers and food availability26; and two of the included studies

used Standard Industrial Classification (SIC) codes25,29, a federally recognized system to

classify stores by industry using a four-digit code. Four articles referred to ambiguous

classification systems, but did not elaborate or explicitly define store types21,22,24,28. Seven

studies did not classify stores or provide store type definitions20,23,27,31,33,35,36.

Store Enumeration

Twelve of the 17 articles20-26, 28-30, 32-34 described their store enumeration methods. Eleven

papers20-22, 24-26, 28,30,32-34 indicated that they had used administrative datasets for store

enumeration, of which four21,22,26,30 also used ground-truthing methods (systematic direct

observation in the field) to corroborate the administrative list of stores. The remaining

paper29 consulted with community organizations for store enumeration.

Survey Adaptations

The majority of included papers 20-29, 32-34, 36 made some type of modification or

adaptation to the original NEMS-S survey. The most common rationale for modification

(nine studies)20-25, 27, 28,33 was to adapt the survey to collect data on a specific food group,

diet type, or nutrient considered to be relevant to the study population. Eight

papers 23,24,26,32,34,36 adapted the survey to make it more appropriate for ethno-cultural

dietary behaviors or to address the local food supply. Three adapted the survey to collect

43

data on a specific store program or policy32,34,36 and two papers used surveys modified for

low-income populations25,29.

Regular and Healthier Foods

Although the original NEMS-S survey collects data on both regular and healthier options,

less than half (seven)25,26,29-32,35 of the included papers reported on both healthy and

unhealthy options. Ten articles reported on healthier foods only20-24,27,28,33,34,36, with three

reporting on fresh fruit and vegetables only23,27,33.

Availability, Price and Quality

Approximately half of the included studies (nine)21, 24, 27,30-35 reported on all three aspects

of the NEMS-S survey. Two reported data on availability and price23,29. Six reported data

on availability only 20,22,25,26,28,36 (Table 5-1).

Availability

All of the included studies reported on food availability. In the original NEMS-S survey,

availability is assessed using three measures: item available in stock (yes/no), number of

varieties available (fruit and vegetables only), and shelf space (milk only). Only three of

the included articles reported all three aspects of availability31,32,34, seven measured stock

and variety22,25,26,28-30,33, and six reported on stock only20,21,23,24,27,36. Variety was measured

and reported in many ways. Six articles used the NEMS-S method (or similar), analyzing

number of varieties using pre-determined tertiles or quartiles25,28-31,33. Two articles used a

similar categorical system but categories were developed based on the variety counts in

44

the sample22,26. Two articles measured raw variety and reported a mean number of variety

of items by store type.32,34

Price

Eleven21,23,24,27,29-35 of the 17 studies reported on price. Of those that reported on price,

two29,31 of the included studies measured price according to NEMS-S protocol; two

studies21,25 modified the pricing method to classify the price of an item as higher or lower

than the median range of prices in a category across all stores and three studies23,32,34

looked at raw or absolute price of an item and one study30 used both raw price and

relative price (healthy vs unhealthy and relative to other store types). Two articles33,35

modified the original NEMS-S survey and provided limited detail on how they distributed

points for price. The final paper27, used a separate tool to measure price.

Quality

Nine (21,24,27,30-35) of the 17 articles reported on the quality of fresh produce. Three

studies27,32,34 analyzed quality independent of a scoring system; all three analyzed quality

by calculating the proportion of items determined to be acceptable using a dichotomized

outcome.

Scoring

Few studies analyzed their data using the original NEMS scoring method. Only four of

the included studies used the NEMS-S composite score29-31,33 (Table 5-1), all of which

also used the NEMS-S sub-scores. Two studies used the availability sub-score only 20,28,

45

one of which considered fat only20. Five studies21,22,24,25,35 used independent scoring

systems (Table 5-1), defined as any scoring system other than the NEMS or significant

adaptations to the NEMS score.

46

Table 5-1 Summary of Review Findings

Rurality Defined

Store Enumeration

Store Types Defined

Reported Measures Score

Gro

und-

Tr

uthi

ng

Seco

ndar

y So

urce

Hea

lthy

Item

s

Unh

ealth

y Ite

ms

Ava

ilabi

lity

Pric

e

Qua

lity

NEM

S-S

Inde

pend

ent

Lasley, E (2008)

! * * ! ! !

Hermstad, A (2010)

! ! !

Gantner, L (2011)

! ! !

! !

Hartley, D (2011)

! ! ! ! ! ! ! !

Hubley, T (2011)

! ! ! ! ! ! !

Johnson, J (2012)

! ! ! ! !

Chau, C (2013)

! ! !

Gantner, L (2013)

! ! ! ! ! !

Pitts, S (2013)

! ! ! ! ! ! !

Vilaro, M (2013)

! ! ! ! ! ! ! ! ! !

Pererira, R (2013)

* * ! ! ! ! ! !

Tisone, C (2014)

! ! ! ! ! ! ! !

Byker-Shanks, C (2015)

! ! ! ! ! ! !

Lu, W (2015)

! ! ! ! ! ! !

Martinez-Donate, A (2015)

* * ! ! ! ! ! !

Olendizki, B (2015)

! ! ! ! ! !

Steeves, E (2015)

! * * ! !

*Insufficient detail provided to draw conclusion

47

5.5 DISCUSSION

The objective of this research was to systematically review the literature to synthesize

how features of the NEMS-S are used to measure the rural consumer food environment,

to examine how the NEMS-S has been adapted in rural food environment settings, and

consider the implications for knowledge users. Six key findings emerged from our

review, which can be broadly categorized as having methodological implications or

implications for reporting and knowledge translation.

Methodological Implications

One of the most commonly-acknowledged barriers to advancing retail food environment

research is the heterogeneity of measures and methodology within the field2,3,7,37,38, and

our findings corroborated this concern. A lack of transparency in the methods used in data

compilation and data classification, and few standardized definitions have been

acknowledged as challenges in synthesizing food environment evidence and drawing

robust conclusions 2,3,7,37,16.

Defining Rurality

Defining rurality in health research is challenging: a wide range of communities and

populations are considered rural on the urban-rural continuum39-42. In the reviewed

studies, classification methods and definitions for what is considered a rural community

varied considerably across studies and included densities, relative proximity, and absolute

populations.

48

Reviews on the most frequently used taxonomies to define urban and rural in US public

health research and policy have been done in the past.39-41 Rural demography is complex

and a single definition cannot justifiably capture all aspects. Dichotomous definitions are

limited in their ability to capture the heterogeneity among degrees of rurality and

remoteness, and evidence suggests that when possible more precise definitions should be

applied.39-41 Rural classification should be applied judiciously, based on the context of the

research question39-41, and ideally with consultation from geographic professionals.41 Our

findings emphasize food environment researchers should explicitly indicate how they

define rurality, accompanied with a clear rationale. When making comparisons across

studies, researchers should be aware of discrepancies in rurality, consider revising

definitions, and avoid aggregating data with dissimilar geographic units.

Store Enumeration

Systematic direct observation in the field (ground-truthing) has been recognized as the

gold standard to enumerate neighbourhood foodscapes.16,43 To save time and resources,

researchers commonly start with secondary analysis of administrative datasets including

online directories, commercial business directories, and government administrative

databases.43 A strength of using the NEMS-S is that completing the on-site survey

provides an opportunity to confirm administrative spatial data. Relying on secondary data

to identify stores for consumer food environment assessments creates the potential for

outlets to be missed during surveying because they were not listed, misclassified, or could

not be found.

49

In this review, only four studies reported verification using ground-truthing; evidence

suggests error is more likely for small or independent stores44,45 and in rural areas.44-46

Secondary sources should be verified with direct field observation when possible, or at

the minimum be combined with additional secondary sources.

Store Classification

Classifying store types can be difficult in rural food environments as the retail presence in

rural communities is often comprised of non-traditional or “hybrid” food outlets.16,47 Ten

of the 17 included studies in this review did not note their classification system or provide

explicit store type definitions; the remaining seven used a variety of standardized and

independently-established criteria to classify stores. A review by Gamba et al.3 reviewed

51 retail food environment articles and found over 32 store types/definitions. An earlier

review by Larson et al48 made similar observations. Methods of classification in the

literature have included store size, annual sales, number of cash registers, number of

employees, variety of foods offered, owner interpretation and federally established

industry classification codes.3,48

Despite efforts within the field, there is yet to be a standardized classification system

recognized as the gold standard. However, the North American Industry Classification

System (NAICS) is among the most widely used classification systems in nutrition

environment literature.6,3,48 NAICS, a successor to the Standardized Industry

Classification (SIC) codes, was designed for statistical purposes and to provide common

definitions and increase comparability across the three countries.49 A potential limitation

50

in using datasets with preassigned NAICS codes is that some investigations have

demonstrated misclassification of NAICS codes in secondary sources which can impact

the sensitivity and positive predictive value of the data source and contribute to both an

under-count and over-count of store types.44,50 Manual verification is recommended.

Implications for Reporting

Unhealthy Food Environments

Results of this synthesis indicate that researchers are predominately using the NEMS-S to

report on healthier food options only, contributing to a notable gap in the literature in

measuring and intervening in consumer exposures to non-nutritious foods. Recent

evidence within community food environment research examining geographic

associations with diet and health outcomes suggests that relative exposures may be more

influential than absolute exposures when exploring food environment associations with

key outcomes such as dietary behaviors51 purchasing patterns52 and weight status53. These

findings have led to recommendations for a wider use of relative measures in food

environment research and the development of interventions to improve the balance

between healthy and unhealthy exposures in retail food environments.51-54 This is also of

importance in considering food environments outside of the United States. For example,

in Canada, there is better evidence for widespread food ‘swamps’ (relative predominance

of unhealthy food sources amplifying material disadvantage) than food deserts.37

51

Quality and Price

A strength of the NEMS-S is that it includes multifaceted measures of food access in the

consumer food environment. Although poor quality and high prices are known barriers to

food access in rural and remote communities55-58, these features are frequently overlooked

in food environment assessment2. Only nine of the 17 included studies in our review

reported on all three aspects of the survey21,24,27,30-35.

Despite the evidence from qualitative and perceived food environment assessment

indicating that quality impacts food accessibility for rural dwellers55-58, little research has

directly assessed food quality in rural retail stores2. Offering and maintaining high quality

supplies of fresh food is often a challenge for small rural stores due to remote location,

access to suppliers, limited turnover, and other infrastructural limitations.59

In the reviewed studies, price was measured and evaluated in various ways. Consumer

food environment researchers should consider their study objectives carefully when

deciding how to operationalize and evaluate price, as well as how the evidence generated

will be used to inform knowledge users, the public, and to design potential interventions.

Analyzing relative price disparities between regular (less healthy) and healthier items

within a store is effective in examining whether the retail food environment offers health-

promoting environmental features and incentivizes consumers to make healthful choices,

but it does not capture overall price nor does it allow for comparisons among stores or

geographic areas. For example, if a rural store sells a loaf of white bread for $6.00 and a

52

loaf of whole grain bread for $6.50 and an urban store sells white bread for $3.50 and

whole grain bread for $4.00, both stores would be scored equally and the overall price

discrepancy between the jurisdictions is not captured. Both relative and absolute measures

are pertinent when it comes to price, so depending on study objectives, researchers may

want to consider using both measures. The median price is also a valuable benchmark for

comparisons between urban and rural areas.

Score

The NEMS-S composite score is easily interpreted and allows for a global assessment of

a store. This is useful for those measuring area-level variations or in intervention studies.

However, without context, scores may have little meaning to knowledge users.18 For

decision-makers at the retail ‘front-line’ such as public health practitioners and retail store

owners, managers and consumers, or for policymakers, it is critical to identify where

change is needed inside the store, and how both incremental and larger changes could be

made. A combination of both composite scores and description is likely the most

appropriate way to identify priority areas for improvement, and to promote multiple kinds

of evidence-informed policy and decision making.

Limitations

This review had a few limitations. First, we focused on NEMS-S and NEMS-CS research

only. As such, the identified gaps that we detected, such as the varied and limited

reporting on unhealthy foods, may not be generalizable to other consumer food

environment literature. However, it is reasonable to posit that since NEMS is considered a

53

gold standard method, the variation may be even greater in other studies. Second, all of

the literature captured in this review was from the United States. Consumer food

environment research is growing worldwide, including rural and remote studies. Given

the wide variability in the quality of rural food environments, it will be important to

capture methodological trends as they emerge in other jurisdictions, through other

synthesis strategies.

5.6 CONCLUSION & RECOMMENDATIONS

This is the first review to exclusively consider rural consumer food environment research

using the NEMS-S measure. Our review demonstrates considerable heterogeneity in the

terminology and taxonomies used to describe and define key constructs of the

methodology including: community classification, store enumeration and store

classification. Our findings also indicate that the majority of NEMS-S research in rural

areas uses an adapted version of the original tool. Many researchers report availability

only and fewer report the price and quality of foods. Most researchers do not report

unhealthy items and few are using the original NEMS-S scoring system.

Consumer food environment researchers should consider their study objectives when

deciding how to classify, define, and operationalize variables of interest. Echoing earlier

reviews, we encourage greater transparency in reporting methodology; and providing

explicit definitions and rationales so that knowledge users may consider the context of the

54

research when interpreting the findings. When making comparisons across studies,

researchers should be aware of discrepancies and avoid aggregating dissimilar methods.

Availability, quality, and price, both relative and absolute, are important features in

measuring the foodscape, especially in rural areas where populations face unique barriers

to food accessibility. Future research should consider elaborating upon all of these

features of the consumer food environment.

In order to inform interventions and influence healthy changes in the retail food

environment, it is essential to produce evidence that is readily applied by decision

makers. A combination of both composite scores and descriptive analysis will allow

researchers to describe the food environment in a format that is meaningful to key

stakeholders, improving their ability to interpret and act on the evidence.

55

REFERENCES

1. Glanz K, Sallis J, Saelens B, Frank L. Healthy nutrition environments: Concepts and measures. Am J Health Promot. 2005 Jun;19(5):330–3.

2. Capsi C, Sorensen G, Subramanian SV, Kawachi I. The local food environment and diet: A systematic review. Health Place. 2012 Sep;18(5):1172–87.

3. Gamba R, Schucjter J, Rutt C, Seto E. Measuring the food environment and its effects on obesity in the United States: a systematic review of methods and results. J Community Health. 2015 Jun 40(3):464–75.

4. Bader M, Ailshire J, Morenoff J, House J. Measurement of the local food environment: a comparison of existing data sources. Am J Epidemiol. 2010 Mar;171(6):609–17.

5. Rose D, Bodor J, Hutchinson P, Swalm C. The importance of a multi-dimensional approach for studying the links between food access and consumption. J Nutr. 2010 Jun;140(6):1170–4.

6. Kelly, Flood, Yeatman. Measuring local food environments: An overview of available methods and measures. Health Place. 2011 Nov;17(6):1284–93.

7. Gustafson A, Hankins S, Jilcott Pitts S. Measures of the consumer food store environment: a systematic review of the evidence 2000-2011. J Community Health. 2012 Aug;37(4):897–911.

8. Kersten E, Laraia B, Kelly M, Adler N, Yen I. Small food stores and availability of nutritious foods: a comparison of database and in-store measures, Northern California, 2009. Prev Chronic Dis. 2012 July;9(127):1037–44.

9. Moore L, Diez Roux A, Franco M. Measuring availability of healthy foods agreement between directly measured and self-reported data. Am J Epidemiol. 2012;175(10):1037–44.

10. Glanz K, Sallis J., Saelens BE, Frank L. Nutrition Environment Measures Survey in stores (NEMS-S): Development and evaluation. Am J Prev Med. 2007 Apr;32(4):282–9.

11. Minaker L, Raine K, Wild T, Nykiforuk C, Thompson M, Frank. Construct validation of 4 food-environment assessment methods: Adapting a multitrait-multimethod matrix approach for environmental measures. Am J Epidemiol. 2014 Feb;179(4):519–28.

56

12. Glanz K, Johnson L, Yaroch A, Phillips M, Ayala G, Davis E. Measures of retail food store environments and sales: Reviews and implications for healthy eating initiatives. J Nutr Educ Behav. 2016 Apr;48(4):280–8.

13. Cavanaugh E, Mallya G, Brensinger C, Tierney A, Glanz K. Nutrition environments in corner stores in Philadelphia. Prev Med. 2013 Feb;56(2):149–51.

14. Honeycutt S, Davis E, Clawson M, Glanz K. Training for and dissemination of the Nutrition Environment Measures Surveys (NEMS). Prev Chronic Dis. 2010 Nov;7(6).

15. Lendarson J, Hansen A, Hartley D. Rural and Remote Food Environments and Obesity. Curr Obes Rep. 2015 Mar 4(1):46–53.

16. Sharkey J. Measuring potential access to food stores and food-service places in rural areas in the U.S. Am J Prev Med. 2009 Apr;36(4):151–5.

17. Minaker L. Retail food environments in Canada: Maximizing the impact of research, policy and practice. Can J Public Health. 2016 Jun;107:5632.

18. NEMS Related Publications [Internet]. Nutrition Environment Measures Survey. Available from: http://www.med.upe nn.edu/nems/docs/NEMSRelatedPubs.7.12.16.pdf

19. Arksey, H. and O’Malley, L. (2005). Scoping studies: Towards a methodological framework. International Journal of Social Research Methodologies, 8(1), 19-32. http://dx.doi.org/10.1080/1364557032000119616 20. Hermstad A, Swan D, Kegler M, Barnette J, Glanz K. Individual and environmental

correlates of dietary fat intake in rural communities: A structural equation model analysis. Soc Sci Med. 2010 Jul;71(1):93–101.

21. Hartley D, Anderson N, Fox K, Lendarson J. How does the rural food environment affect rural childhood obesity?. Child Obes. 2011 Dec;7(6):450–61.

22. Gantner L, Olsen C, Frongillo E. Relationship of food availability and accessibility to women’s body weights in rural upstate New York. J Hunger Environ Nutr. 2013 Nov; 8(4):490–505.

23. Lasley EC., Litchfield R. Fresh produce in rural Iowa: Availability and Accessibility. J Hunger Environ Nutr. 2008 Oct;2(2–3):5–17.

24. Hubley T. Assessing the proximity of healthy food options and food deserts in a rural area In Maine. Applied Geography. 2011Oct;31(4):1224–31.

57

25. Olendzki B, Procter-Gray E, Wedick N, Patil V, Zheng H, Kane K, et al. Disparities in access to healthy and unhealthy foods in central Massachusetts: implications for public health policy. J Am Coll Nutr. 2015;34(2):150–8.

26. Gantner L, Olson C, Frongillo E. Prevalence of nontraditional food stores and distance to healthy foods in a rural food environment. J Hunger Environ Nutr. 2011 Aug; 6: 279–93.

27. Johnson J., Nobmann ED, Asay E. Factors related to fruit, vegetable and traditional food consumption which may affect health among Alaska Native People in Western Alaska. Int J Circumpolar Health. 2012 Mar; 71(0):183–8.

28. Chau C, Zoellner J, Hill J. Availability of Healthy Food: Does Block Group Race and Income Matter? J Hunger Environ Nutr. 2013 Mar; 8(1):22–38.

29. Pitts S, Lawton K, McGuit J, Wall-Bassett E, Morgan J, Laska M, et al. Formative evaluation for a healthy corner store initiative in Pitt County, North Carolina: assessing the rural food environment, part 1. Prev Chronic Dis. 2010 Jul;10.

30. Vilaro M, Barnett T. The rural food environment: A survey of food price, availability, and quality in a rural Florida community. Food and Public Health. 2013;3(3):111–8.

31. Pereira R, Sidebottom A, Boucher J, Lindberg R, Werner R. Assessing the food environment of a rural community: baseline findings from the heart of New Ulm project, Nutrition Environment Measures Survey in rural communities Minnesota, 2010-2011. Prev Chronic Dis. 2014 Mar;6(11).

32. Tisone C, Guerra S, Lu W, McKyer E, Ory M, Dowdy D, et al. Food-shopping environment disparities in Texas WIC vendors: a pilot study. J Health Behav. 2014 Sep;38(5):726–36.

33. Byker Shanks C, Ahmed S, Smith T, Houghtaling B, Jenkins M, Margetts M, et al. Availability, Price, and Quality of Fruits and Vegetables in 12 Rural Montana Counties, 2014. Prev Chronic Dis. 2014 Sep;12(E128).

34. Lu W, McKyer E, Dowdy D, Evans A, Ory M, Hoelscher D, et al. Evaluating the Influence of the Revised Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) Food Allocation Package on Healthy Food Availability, Accessibility, and Affordability in Texas. J Acad Nutr Diet. 2016 Feb;116(2):292–301.

35. Martinez-Donate A, Riggall A, Meninen A, Malecki K, Escaron A, Hall B, et al. Evaluation of a pilot healthy eating intervention in restaurants and food stores of a rural community: a randomized community trial. BMC Public Health. 2015 Feb;12(15):136.

58

36. Steeves EA, Penniston E, Rowan M, Steeves J, Gittelsohn J. A Rural Small Food Store Pilot Intervention Creates Trends Toward Improved Healthy Food Availability. J Hunger Environ Nutr. 2015 Jun;10(2):259–70.

37. Minaker L, Shuh A, Olstad D, Engler-Stringer R, Black J, Mah C. Retail food environments in Canada: A scoping review. Can J Public Health. 2016 Jun; 107(1):ES4-ES13.

38. Lucan S. Concerning Limitations of Food-Environment Research: A Narrative Review and Commentary Framed around Obesity and Diet-Related Diseases in Youth. J Acad Nutr Diet. 2015 Feb;115(2):205–12.

39. Vanderboom C, Madigan E. Federal definitions of rurality and the impact on nursing research. Res Nurs Health. 2007 Apr;30(2):162–75.

40. Hall S, Kaufman J, Ricketts T. Defining urban and rural areas in U.S. epidemiologic studies. 2006 Mar;83(2):162–75.

41. Hart L, Larson E, Lishner D. Rural definitions for health policy and research. m J Public Health. 2005 Jul;95(7):1149–55.

42. Lyseen A, Hansen H, Harder H, Jensen A, Mikkelsen B. Defining Neighbourhoods as a Measure of Exposure to the Food Environment. Int J Environ Res Public Health. 2015 Jul;12(7).

43. Fleischhacker S, Evenson K, Sharkey J, Pitts S, Rodriguez D. Validity of secondary retail food outlet data: a systematic review. Am J Prev Med. 2013 Oct;45(4):462–73.

44. Liese A, Colabianchi N, Lamichane A, Barnes T, Hibbert J, Porter D, et al. Validation of 3 food outlet databases: completeness and geospatial accuracy in rural and urban food environments. Am J Epidemiol. 2010 Dec; 172(11):1324–33.

45. Sharkey J, Horel S. Neighborhood socioeconomic deprivation and minority composition are associated with better potential spatial access to the ground-truthed food environment in a large rural area. J Nutr. 2008 Mar;12(7):8504–25.

46. Fleischhacker S, Rodriguez D, Evenson K, Henley A, Gizlice Z, Soto D, et al. Evidence for validity of five secondary data sources for enumerating retail food outlets in seven American Indian communities in North Carolina. Int J Behav Nutr Phys Act. 2012 Nov. 22(9):137.

47. Bustillos B, Sharkey J, Anding J, McIntosh A. Availability of more healthful food alternatives in traditional, convenience, and nontraditional types of food stores in two rural Texas counties. J Am Diet Assoc. 2009 May;109(5):883–9.

59

48. Larson N, Story M, Nelson M. Neighborhood environments: Disparities in access to healthy foods in the U.S. Am J Prev Med. Nov 2008;36(1).

49. North American Industry Classification System (NAICS) Canada [Internet]. Ottawa, Ontario: Statistics Agency of Canada; 2016. Available from: http://www.statcan.gc.ca/pub/12-501-x/12-501-x2016001-eng.pdf

50. Powell L, Han E, Zenk S, Khan T, Quinn C, Gibbs K, et al. Field validation of secondary commercial data sources on the retail food outlet environment in the U.S. Health Place. 2011 Sep;17(5):1122–31.

51. Clary C, Ramos Y, Shareck M, Kestens Y. Should we use absolute or relative measures when assessing foodscape exposure in relation to fruit and vegetable intake? Evidence from a wide-scale Canadian study. PrevMed. 2015 Feb;71:83–7.

52. Mason K, Bentley R, Kavanagh A. Fruit and vegetable purchasing and the relative density of healthy and unhealthy food stores: evidence from an Australian multilevel study. 2013 Mar;67(3):231.

53. Polsky J, Moineddin R, Glazier R, Dunn J, Booth G. Relative and absolute availability of fast-food restaurants in relation to the development of diabetes: A population-based cohort study. Can J Public Health. 2016 Jun;107:5312.

54. Zenk S, Powell L, Rimkus L, Isgor Z, Barker D, Ohri-Vachaspati P, et al. Relative and Absolute Availability of Healthier Food and Beverage Alternatives Across Communities in the United States. Am J Public Health. 2014 Nov;104(11):2170–8.

55. Smith C, Morton L. Rural food deserts: low-income perspectives on food access in Minnesota and Iowa. J Nutr Educ Behav. 2009 Jun;41(3):176–87.

56. Dean W, Sharkey J. Rural and urban differences in the associations between characteristics of the community food environment and fruit and vegetable intake. J Nutr Educ Behav. 2011 Nov;43(6):426–33.

57. Maley M, Warren B, Devine C. Perceptions of the environment for eating and exercise in a rural community. J Nutr Educ Behav. 2010 May;42(3):185–91.

58. Ladouceur L, Hill F, Indian Northern Affairs Canada. Results of the survey on food quality in six isolated communities in Labrador, March 2001 [Internet]. Ottawa Ontario:Government of Canada; 2002. Available from: http://publications.gc.ca/collections/ Collection/R2-186-2001E.pdf

59. Office of Nutrition Policy and Promotion, Health Canada. Working with Grocers to Support Healthy Eating [Internet]. Ottawa Ontario: Government of Canada; Available from: https://www.canada.ca/en/health-canada/services/food-nutrition /healthy-eating/nutrition-policy-reports/working-grocers-support-healthy-eating.html

60

CHAPTER 6: The Nutrition Environment Measures Survey Newfoundland and

Labrador: An Exploratory Analysis into the Significance of Assessing the Retail Food

Environment Using Healthy Food Only

61

AUTHORSHIP STATEMENT

TITLE: The Nutrition Environment Measures Survey Newfoundland and Labrador: An Exploratory Analysis into the Significance of Assessing the Retail Food Environment Using Healthy Food Only

AUTHORS: 1. Rebecca Harris (Lead Author), MSc Student, Community Health and Humanities, Memorial University of Newfoundland and Labrador 2. Dr. Catherine Mah, Associate Professor, School of Health Administration, Dalhousie University 3. Dr. James Valcour, Associate Professor, Faculty of Medicine, Memorial University of Newfoundland and Labrador 4. Dr. Leia Minaker, Assistant Professor, School of Planning, University of Waterloo

Contributions

RH proposed the concept for this manuscript. RH and CM devised a design and

implementation plan. RH conduced the literature review. Data for this manuscript was

initially collected by the Healthy Corner Store Newfoundland and Labrador research team.

RH prepared the data by creating the revised NEMS-S scoring template, CM and LM

reviewed template. Next, RH exported, cleaned, recoded and scored data according to

scoring template. RH devised analysis plan with consultation from CM with additional

input from JV and LM. RH analyzed and interpreted data.

All drafts of the manuscript (including all sections) were written by RH. Initial and

intermediate versions were sent to CM and JV for edits and suggested revisions. Final

versions were sent to LM for edits.

62

6.1 ABSTRACT

The objective of this study was to explore the significance of assessing rural consumer

food environments using healthy food availability only. To achieve our objective, we

assessed healthy, regular (less healthy) and total (both healthy and regular) food

availability in a census of supermarkets and convenience stores in rural Newfoundland

and Labrador.

Using a scored checklist measure, supermarkets earned greater scores for the availability

of healthy, regular and total food availability. Both convenience stores and supermarkets

offered most regular food and beverage options while there was greater variability among

the availability of healthier items.

However, our results demonstrated a strong and positive correlation between healthy food

availability scores and regular food availability scores, supporting theories that high

exposure to healthy items is often accompanied by a comparably high exposure to

unhealthy items. Although healthy food availability may be an appealing measure to

capture the variability among stores, it may not represent the overall consumer food

environment and the exposures people experience while shopping.

63

6.2 BACKGROUND

Over several decades, public health efforts to prevent and reverse the growing burden of

obesity and diet-related diseases1 have progressively taken a socio-ecological perspective.

In order to observe change in individual behaviors, there must be supportive built,

socioeconomic, and policy environments, in which to make healthy choices.2,3,4

Considerable attention has been given to the retail food environment and its potential

influence on food choice and dietary behavior. The retail food environment, as defined by

Glanz and colleagues, can be organized into 2 main components: the community nutrition

environment defined as the proximity or density of varying types of food outlets within a

defined area and the consumer nutrition environment, the availability, price and quality of

healthy food and beverage within a retail food outlet.2

A growing body of literature has illustrated both the significance and variability of the

availability, affordability and quality of food and beverage within wide-ranging

geographic and socioeconomic environments.5,6,7,8,9 Among the most vulnerable

environments are rural areas, where evidence has illustrated access to full service

supermarkets and grocery stores is significantly limited and there is a higher proportion of

small convenience stores compared to urban communities.5,10,11,12 This is concerning

because of evidence that indicates larger supermarkets and grocery stores have a wider

variety of affordable nutrient dense foods compared to smaller convenience and corner

stores where there is a higher availability of less nutritious, energy dense, snack

64

foods.5,10,11,12 When healthier items are available in small stores, they tend to be of a

higher price and lower quality.5,10,11,12

In Canada, although the population is urbanizing, there is considerable variation in the

proportion of population living in rural and remote areas between jurisdictions. For

example, in the province of Newfoundland and Labrador (NL), 40% of the population

continue to reside in rural or remote areas.13 Consistent with other rural food environment

literature, food access in many NL communities is limited; industry data as well as a

recent census of rural stores by our team, indicates that about 7 in 10 stores in rural

communities in NL are convenience stores, suggesting convenience stores play a

substantial role in food choices in rural NL.14

Diet quality and diet-related health status is also of concern in NL. Only 26% of

individuals 12 and older consume at least 5 servings of vegetables and fruit per day

compared to the national average of 40%.2 NL leads the nation in the prevalence of

overweight and obese individuals, with over two thirds of the provincial population

classified as overweight or obese.15 These estimates exceed those in all other provinces

and territories. NL also faces a high burden of other diet-related diseases including type 2

diabetes, hypertension, cardiovascular disease and osteoarthritis.15,16

Early food environment research in rural areas has drawn the interest of researchers,

policy makers and other key stakeholders and has demonstrated an urgency for action.

However, in order to encourage action and guide priority-setting for policy and decision

65

making, it is essential for researchers to capture accurate measures of the rural foodscape.

Much of retail food environment research has been dedicated to measuring spatial access

to healthy food and the concept of food deserts5,6,7,8, which are communities or other

geographic areas in which inadequate access to healthy foods exacerbates economic

disadvantage.17 However, in more recent work, researchers have begun to acknowledge

and measure the increasing accessibility to non-nutritious food and beverages.

In 2009 Rose and colleagues introduced the concept of the food swamp, defined as areas

with relatively few healthy options or where “large relative amounts of energy-dense

snack foods, inundate healthy food options”.18 This has encouraged researchers to

become more attentive to the relative availability of healthy food compared to unhealthy

foods, in contrast to healthy food availability alone. 19,20,21,22,23,24,25,26,27,28,29,30 Still-

advances in relative food environment assessment have largely been conducted in the

context of community-level food environment (the relative density and proximity of food

retailers that have been generally accepted as healthy or unhealthy31) and few researchers

have applied concepts of relative food availability to in-store assessment within the

consumer food environment.20,23,25 26,27,28,29,30

Consumer food environment assessment usually involves direct observation of the retail

food environment and has been considered a more sophisticated and robust measure of

the food environment compared to community-level assessment.32,33 Direct observation

can include checklists, variety counts, display counts or shelf space measures of the foods

and beverages within stores. Although some of these approaches are relatively resource

66

intensive, checklist measures of select indicator items can be an efficient and effective

method for those with limited resources.

A recent review on food environment assessment indicated that the use of checklist

measures to audit stores has more than doubled since the publication of a previous review

in 2007, and they are now the most frequently used tools to evaluate the retail food

environment.4 However, despite emerging literature on the importance of relative food

environment measures, many of those using checklist measures are assessing healthy food

availability only. Our review of rural food environment research using the Nutrition

Environment Measures Survey for Stores (a checklist and gold standard in consumer food

environment tools assessing the availability, price and quality of ‘healthier’ and ‘regular’

(less healthy) options of food and beverages) demonstrated that many researchers (59%)

reported on measures of healthier food and beverages only and not “regular options”

(typically the same food or beverage product with additional fat, sugar, or salt) to measure

store healthfulness. 34

This is a substantial research gap, as researchers who have implemented relative food

environment measures using checklists have demonstrated a strong relationship between

healthy food availability and unhealthy food availability, suggesting that stores that offer

a high variety of healthy foods may also offer a high variety of unhealthy foods.28,30

Research using shelf space or display count measures have shown that many

supermarkets, often coded as healthy food sources in community food environment

67

research, dedicate a greater volume of shelf space to energy dense snack foods compared

to fruits, vegetables and other nutritious food.23,29,25,27

These findings suggest the relative exposure of healthy and unhealthy food may be more

influential than absolute exposures when exploring food environment associations with

key outcomes of interest such as dietary behaviors21, purchasing patterns22, and weight

status19. These findings raise questions regarding measuring accessibility to healthy foods

only and the adequacy of this method as an indicator of store healthfulness and merits

further investigation.

6.3 OBJECTIVES AND RESEARCH DESIGN

The objective of this study is to use the NEMS-S checklist measure to assess healthy food

availability, regular (less healthy) food availability and total food availability (both

healthy and regular food availability) in a census of retail food stores in eastern

Newfoundland, and to compare and critically discuss measurement outcomes. The

ultimate goal of our research was to gain a greater understanding of the significance and

implications of assessing the consumer food environment using healthy food only, most

specifically through the NEMS-S checklist measures which is among the most widely

used consumer food environment tools.

Data for this project was provided by the Healthy Corner Stores Newfoundland and

Labrador (HCSNL) project, a pilot program to define and test the potential for healthier

68

foods in Newfoundland convenience stores. In HCSNL, researchers used an adapted

version of the Nutrition Environment Measures Survey (NEMS-NL) to collect data on the

availability (including variety), quality and price of healthy foods in a census (n=78) of

rural food stores across the Avalon Peninsula in Eastern NL.

6.4 METHODS

Sample

HCSNL conducted NEMS-NL in census of rural food stores across the Avalon Peninsula,

the most populous region of the province. Although over 50% of the provincial

population live within the Avalon, 75% of residents on the Avalon live in the capital city,

St. John’s, and the surrounding metropolitan area, leaving a rural population density of

only 7.7 people per square mile.

“Rural” was defined as the area outside the metropolitan region within the Avalon

Peninsula. Stores were further classified by rurality into one of six categories according to

the NL-Accessibility Remoteness index: Highly accessible, Accessible, Somewhat

Accessible, Moderately Remote, Remote and Extremely Remote.35 The NL-Accessibility

Remotes index classifies community remoteness by level of accessibility using a complex

algorithm that takes into account population size and travel time to various community

services (i.e, public schools, primary health care). “Food Store” was defined as any

retailer classified into one of three North American Industry Classification System codes:

Supermarkets and Other Grocery, Convenience Stores and Gas stations with Convenience

69

Stores”.36 More detailed definitions for remoteness levels and store types can be found in

in Appendix 6-A, as well as from their respective sources.35,36

The final census assessed by HSCNL was n=78, including 17 grocery stores, 48

convenience stores and 18 gas stations with convenience stores. For the purposes of the

current analysis we collapsed both convenience store types (both with and without gas

stations) into a single category. Eight stores were classified as highly accessible, 44 were

accessible, 17 were somewhat accessible and 9 were classified as moderately remote. No

stores were classified as remote or extremely remote

Data Collection Tool

The Nutrition Environment Measures Survey for Stores (NEMS-S) is an environmental

survey to facilitate direct, observational measurement of food and beverage availability,

pricing, and quality within retail food stores. NEMS-S collects data on both “healthier”

and “regular” alternatives for food and beverages that are part of an average American

diet and those recommended for healthy eating. The NEMS-S classifies an item as a

“healthier” or “regular” option in relative terms within individual measures (i.e. white

bread and whole wheat bread, full fat chips and baked chips).37,38 This model allows for

researchers to measure the consumers’ opportunity to make healthier choices for items

that are part of their usual diet. NEMS-S is accompanied by a scoring system in which

points are awarded for having healthier items available at a lower prices and acceptable

quality. Sub-scores for availability, price and quality can be summed for a total score

intended to represent overall store healthfulness.37,38

70

NEMS-S was adapted for use in Toronto (ToNEMS-S) by Lo and colleagues to reflect

local diet and consumer patterns as well as Canada’s Food Guide dietary

recommendations.39 The HCSNL researchers collected NEMS-NL data using a modified

version of the ToNEMS-S adapted to reflect the dietary pattern and food availability in

NL. The finalized NEMS-S-NL tool collected information on the availability, price and

quality of 98 food and beverage items (including both healthy and regular versions)

organized into 14 categories. Further details of the NEMS-NL tool and collection for the

HSCNL project have been described elsewhere.14

Data Preparation

To investigate the significance of evaluating store healthfulness using healthy food

availability only, we modified the original NEMS-S scoring system to develop a “healthy

food availability” score, a “regular food availability” score and a “total food availability”

score.37,38

In the NEMS-S scoring system, items classified as healthier and items classified as

regular are not equally nutritious or non-nutritious but this varying degree of

healthfulness is not reflected in point distribution. For example, under the healthy

category, stores are awarded 2 points for the availability of whole wheat bread but are

also awarded 2 points for the availability of baked chips.

71

Recognizing the limitations of classifying foods as healthy or regular using dichotomized

classification, we further classified items by healthfulness using a three-tiered system,

similar to the approach taken in school nutrition or traffic light guidelines. This

classification system was guided by both Canada’s Food Guide and the National

Document for the Development of Nutrient Criteria for Foods and Beverages in

Schools.40 Healthy items were classified as healthiest, less healthy or least healthy and

regular items were classified as least unhealthy, less unhealthy or most unhealthy. Points

were distributed according to healthfulness classification and summed to give a healthy

food availability score and a regular food availability score. Total food availability score

was derived by subtracting the retailer’s regular food availability score from the healthy

food availability score. Scoring templates can be found in Appendix 6B.

Data Analysis

The NEMS-NL scores were analyzed using SPSS Statistics (version 24) software

package. We calculated mean healthy scores, regular scores and total scores, then

compared means across rurality and store type.

Mean healthy scores, regular scores and total scores by level of remoteness and store type

were compared using ANOVA and independent sample t tests. Correlations among scores

were tested using Spearman’s rank correlation test.

We also conducted a series of independent sample t tests to analyze differences in scores

for individual measures (i.e. fruit, meat, cereal) among store types to provide a more

72

detailed description of the variability among stores. By evaluating individual measures,

we can provide data that is more meaningful to key stakeholders, policy makers and store

owners. For example, our previous work has suggested that fresh and frozen produce

including fruits, vegetables and meat are limited in small and rural stores14 by reporting

on individual measures we can identify these potential gaps and other priority areas where

change is needed.

6.5 RESULTS

A total of 78 stores were surveyed (n=61 convenience stores, n=17 supermarkets). Over

half of stores were classified as accessible (n=44), followed by somewhat accessible

(n=17), moderately remote (n=9) and highly accessible (n=9). No stores were classified

as remote or extremely remote.

Remoteness

Mean healthy, regular and total scores did not differ significantly by level of store

remoteness (Table 6-1).

73

Table 6-1 Mean scores by remoteness

Store Type

Healthier, Regular & Total Availability Scores

Supermarkets scored significantly higher than convenience stores in both healthy food

availability and total food availability (p<0.01). However, they also scored significantly

higher for availability of regular food items, earning a mean score of 40.3 out of a

possible 47 points.

Table 6-2. Mean scores by store type

* Significant at p <0.01

Individual Measure Availability Score

In individual healthy measures, supermarkets scored significantly higher for the

availability of most items (Appendix 6C). Many of the remaining items (in which there

Remoteness N Healthy Score (SD)

F=0.01 p=0.99

Regular Score (SD)

F=0.11 p=0.96

Total (SD)

F=0.16 p=0.92

Highly Accessible 8 52.4 (23.4) 32.8 (11.6) 19.6 (12.8)

Accessible 44 51.7 (21.6) 34.66 (11.4) 17.0 (12.6)

Somewhat Accessible

17 52.0 (18.1) 35.2 (10.4) 16.8 (10.4)

Moderately Remote 9 51.1 (13.3) 35.22 (6.9) 15.9 (8.0)

Store Type N Healthy Score (SD)

Regular Score (SD) Total Score (SD)

Supermarket

17 68.1* (19.9) 40.3* (10.7) 27.8* (11.7)

Convenience Store

61 47.2 (17.5) 33.1 (10.2) 14.1 (9.7)

74

were no statistical differences in availability scores), could be considered “staple” items

and were readily available in both store types. These items include: low fat milk, eggs,

healthier cheese, whole wheat bread, healthier baked goods, meat alternatives, healthier

chips, healthier beverages and low sugar cereals.

In contrast, in the individual regular measures, there was little difference in availability of

regular food and beverages among supermarkets and convenience stores. Supermarkets

earned comparable scores to convenience stores in almost all items, with both store types

scoring highly for the availability of most items. Processed meat, processed cheese and

sugary cereals were the only items in which supermarkets scored significantly higher than

convenience stores. Individual measure scores by store type can be found in Appendix

6C.

Correlations

Correlation tests demonstrated a positive correlation (0.88) between healthy and regular

scores at p<0.01 indicating that higher healthy food availability scores are often

complemented by higher regular food availability scores.

6.6 DISCUSSION

This study found that there were no significant differences in availability of healthier or

regular items among stores at different degrees of rurality; this could be a result of the

lack of data points at the higher end of the remoteness index (remote and extremely

75

remote) and the high proportion of stores classified as accessible. However, in terms of

store type, consistent with findings from similar work, our results indicated that

supermarkets had a higher availability of both healthy and regular food and beverage

compared to convenience stores.28 We also observed a strong and positive relationship

between healthy food availability scores and regular food availability scores, as

demonstrated through the correlation statistic (0.88). These findings support those

observed in similar studies; Olendski and colleagues found that healthy food availability

was highly correlated with unhealthy food with a correlation coefficient of 0.90. 28 In

other words, stores that are able to stock more food, do so over a wide range of product

lines, and thus supply a higher quantity of both heathier and less healthy foods. Together,

these results illustrate that it may be presumptuous to define a retailer as a “healthy food

source” based on the assessment of healthy food availability alone.

However, there are aspects of our analysis that demonstrate why it may be appealing for

some researchers to use healthy food availability as a proxy for total (or relative) food

availability when ranking or comparing retailers. When comparing total food availability

(regular food availability scores subtracted from healthy food availability scores) we

observed that despite having a higher regular food availability score, supermarkets

maintained a higher total food availability scores. It is possible that these results can be

explained in part by the pervasive availability of regular food items and a more scarce and

dispersed availability of healthy food. Although supermarkets earned a higher overall

regular food availability score, the point discrepancy was modest (8 points) and observing

individual food categories it was evident that almost all stores, both supermarkets and

76

convenience stores, offered most regular items. There were only three individual regular

categories (processed meat, processed cheese and sugary cereals) in which supermarkets

earned higher scores than convenience stores; in the remaining 9 categories, we did not

detect any difference among convenience stores and supermarkets. Conversely,

examining healthy food availability, the magnitude of difference between supermarkets

and convenience stores was greater; supermarkets scored 23 points higher in overall

healthy food availability and earned higher scores in 10 out of the 19 individual healthy

categories compared to convenience stores.

Although few have reported on overall food availability using a checklist measure, we are

not the first to observe this tendency. 28,30 These findings echo those observed in a similar

study by Zenk and colleagues who examined the absolute and relative availability of 14

indicator items (7 healthy and 7 unhealthy) using a checklist measure.30 Consistent with

our findings, absolute healthy, unhealthy and relative food availability was higher in

supermarkets compared to limited service stores. Others have also observed a greater

variability in healthy food availability while the variability in less healthy food was more

narrow. Oldenski et al. used both healthy and unhealthy food availability indices (scores)

to evaluate healthfulness in small and large stores. In this case, healthy food availability

scores were almost four times higher in large stores compared to small stores (30.1 vs

7.0) while the difference in unhealthy food availability was only half that (13.0 vs 26). 28

A common argument for the assessment of unhealthy food and beverages in stores is the

concept of the food swamp and the widespread availability of less nutritious items,

77

however, the predominance of unhealthy foods could also be a relevant counter-argument

for the assessment of healthy foods only. Because less nutritious food such as chips, soft

drinks and other energy dense snack foods are so readily available in most stores, it is not

likely that will we gain much insight or observe much variability in incremental

differences in store healthfulness through assessing the availability of these items with a

checklist. For those surveying a large sample of stores and/or with limited time or

resources, it may not be efficient to conduct consumer food environment measures or

collect data on both healthy and regular options, especially considering the large number

of varieties of snack foods and sugar sweetened beverages. Collectively, our results and

literature with similar findings, demonstrates why it might be appealing to assess healthy

items only when evaluating store healthfulness, specifically when ranking or comparing a

group of stores using a checklist measure.

However, our research has illustrated that defining stores as “healthy” based on healthy

food availability or total food availability is inadequate without reporting on potential

unhealthy food exposure. As demonstrated through the positive correlation between

healthy and regular food exposure, a store with a higher availability of healthy items

cannot be presumed to be an overall” healthy” retailer. Using language such as “greater

exposure to healthy foods” or “healthier” may be a more appropriate description for

stores that have a higher availability of healthy foods.

It should be emphasized that the intention of these conclusions is not to suggest unhealthy

food is an irrelevant component of food environment research. In fact, our study

78

acknowledges that unhealthy food and beverages have become the most dominant

component of the food environment- so much so that the availability of these items is

often implied or assumed through existing measurement methods. Our research has

demonstrated that rural communities face similar challenges to their urban counterparts in

regards to disproportionate and excessive exposure to unhealthy foods and the need to

refine the sophistication of in-store assessment to contextualize food environment

exposures.

These results also have implications for the design of healthier retail food environment

interventions-particularly in rural communities. To date, the majority of food environment

interventions have been aimed at increasing the availability of healthy foods in

convenience stores or improving geographic accessibility to grocery stores and

supermarkets.41,42,43 However, the effectiveness of these strategies have been mixed;

where interventions have been effective in improving accessibility to healthy foods,

evidence for improvements in health outcomes have often been lacking41,42, suggesting

increasing the accessibility of healthy items is necessary, but not sufficient to make a

significant impact on purchasing and diets. It is perhaps not until effort is made to

intervene in the overwhelming presence of non-nutritious foods that we can expect to

observe any substantial variability in the availability of these items in the retail food

environment.

79

6.7 STRENGTHS & LIMITATIONS

A strength of this work is that we have refined our understanding of scoring within store

audit methods and explored some of the assumptions within scoring methods related to

the NEMS-S survey. In terms of limitations, the results are based on data collecting using

the NEMS-S method in a rural area of Canada, and given the heterogeneity of rural

environments may not be generalizable to other jurisdictions, or to those who conduct

complete store inventory including extensive variety counts and shelf space measurement.

In this case, more variability may be captured within stores. This research was

exploratory work; we did not conduct sensitivity analysis of the healthy, regular or total

availability score, and this may be considered for future research.

6.8 CONCLUSIONS

In this census of rural stores in Eastern NL, Canada, both supermarkets and convenience

stores offered most regular food and beverage options while the availability of healthier

items differed between supermarkets and convenience stores with a greater magnitude

and were less likely to be available in convenience stores. Although healthy food

availability and total food availability are appealing to use as measures in capturing the

disparities among stores, they may not represent the overall consumer food environment

and the exposures people experience while shopping.

80

More investigation is needed to strengthen the measurement of the variability among

consumer food environments. Regardless of store type, less nutritious foods are readily

available to consumers and indeed a meaningful and prevailing component of the retail

food environment. When possible, policymakers and other key stakeholders should

expand retail food environment interventions to address exposure to both healthy and

unhealthy food and beverages.

81

REFERENCES

1. Kassebaum NJ, Barber RM, Bhutta Z, Dandona L, Gething PW, Hay SI, et al. Global, regional, and national levels of maternal mortality, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet. 2016;388(10053):1775–812.

2. Glanz K, Sallis J, Saelens B, Frank L. Healthy nutrition environments: Concepts and measures. Am J Health Promot. 2005 Jun;19(5):330–3.

3. Story M, Kaphingst K, Robinson-O’Brien R, Glanz K. Creating healthy food and eating environments: Policy and environmental approaches. AnnuRevPublic Health. 2008;29:253–72.

4. Lytle L, Sokol R. Measures of the food environment: A systematic review of the field, 2007-2015. Health Place. 2017 Mar;44:18–34.

5. Larson N, Story M, Nelson M. Neighborhood environments: Disparities in acess to healthy foods in the U.S. Am J Prev Med. 2008 Jan;36(1).

6. Walker R, Keane C, Burke J. Disparities and access to healthy food in the United States: A review of food deserts literature. Health Place. 2010 Sep;16(5):876–84.

7. Health Canada. Measuring the food environment in Canada [Internet]. Ottawa Ontario; 2013 Oct. Available from: http://publications.gc.ca/collections/collection_2013/sc-hc/H164-155-2012-eng.pdf

8. Minaker L, Shuh A, Olstad D, Engler-Stringer R, Black J, Mah C. Retail food environments in Canada: A scoping review. Can J Public Health. 2016 Jun;107(1):5344.

9. Hilmers A, Hilmers D, Dave J. Neighborhood disparities in access to healthy foods and their effects on environmental justice. Am J Public Health. 102(9):1664–54.

10. Liese A, Weis K, Pluto D, Smith E, Lawson A. Food store types, availability and costs of food in a rural environment. J Am Diet Assoc. 2007 Nov;107(11):1916–23.

11. Lendarson J, Hansen A, Hartley D. Rural and Remote Food Environments and Obesity. Curr Obes Rep. 2015 Mar;4(1):46–53.

12. Jithitikulchai T, Dean W, Sharkey J. Variations in the availability and price of healthier food options by store type and urban-rural setting. 2012 Dec;7(4):381–400.

82

13. Statistics Canada. Population, urban and rural by province and territory (Newfoundland and Labrador) [Internet]. 2012. Available from: https://www12.statcan.gc.ca/census-recensement/2011/ref/dict/geo042-eng.cfm

14. Authors. BLINDED FOR REVIEW {IN PRESS-UNDER REV}.

15. Statistics Canada. Canadian community healthy survey-annual component summary tables [Internet]. Statistics Canada. 2015. Available from: http://www5.statcan.gc.ca/COR-COR/COR-COR/objList?lang=eng&srcObjType=SDDS&srcObjId=3226&tgtObjType=CST

16. The Cost of Diabetes in Newfoundland and Labrador [Internet]. Canadian Diabetes Association; 2010. Available from: https://www.diabetes.ca/CDA/media/documents/publications-and-newsletters/advocacy-reports/cost-of-diabetes-in-newfoundland-labrador.pdf

17. Gallagher M. USDA Defines Food Desert. Nutrition Digest Published by the American Nutrition Association [Internet]. Vol 38, No.2. Available from: http://americannutritionassociation.org/newsletter/usda-defines-food-deserts

18. Rose D, Bodor N, Swalm C, Rice J, Farley T, Hutchinson P. Deserts in New Orleans? Illustrations of Urban Food Access and Implications for Policy [Internet]. University of Michigan National Poverty Center/USDA Economic Research Service Research; 2009. Available from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.189.2333&rep=rep1&type=pdf

19. Polsky J, Moineddin R, Glazier R, Dunn J, Booth G. Relative and absolute availability of fast-food restaurants in relation to the development of diabetes: A population-based cohort study. Can J Public Health. 2016 Jun;107:5312.

20. Bodor N, Rice J, Farley T, Swalm C, Rose D. Disparities in food access: Does aggregate availability of key foods from other stores offset the relative lack of supermarkets in African-American neighborhoods? PrevMed. 2010 Jul;

21. Clary C, Ramos Y, Shareck M, Kestens Y. Should we use absolute or relative measures when assessing foodscape exposure in relation to fruit and vegetable intake? Evidence from a wide-scale Canadian study. PrevMed. 2015 Feb;71:83–7.

22. Mason K, Bentley R, Kavanagh A. Fruit and vegetable purchasing and the relative density of healthy and unhealthy food stores: evidence from an Australian multilevel study. 2013 Mar;67(3):231.

23. Farley T, Rice J, Bodor N, Cohen D, Bluthenthal R, Rose D. Measuring the food environment: shelf space of fruits, vegetables, and snack foods in stores. J Urban Health. 2009 Sep;86(5):672–82.

83

24. Mercille G, Lucie R, Lise G, Kestens Y, Payette H, Daniel M. Comparison of Two Indices of Availability of Fruits/ Vegetable and Fast Food Outlets. J Urban Health. 2013 Apr;90(2).

25. Miller C, Bodor N, Rose D. Measuring the Food Environment: A Systematic Technique for Characterizing Food Stores Using Display Counts. J Environ Public Health. 2012;2012.

26. Hutchinson P, Bodor N, Swalm C, Rice J, Rose D. Neighbourhood food environments and obesity in southeast Louisiana. Health Place. 2012;18(4):854–60.

27. Vandevijvere S, Mackenzie T, Mhurchu C. Indicators of the relative availability of healthy versus unhealthy foods in supermarkets: A vailidation study. Int J Behav Nutr Phys Act. 2017;14(53).

28. Olendzki B, Procter-Gray E, Wedick N, Patil V, Zheng H, Kane K, et al. Disparities in access to healthy and unhealthy foods in central Massachusetts: implications for public health policy. J Am Coll Nutr. 2015;34(2):150–8.

29. Rose D, Hutchinson P, Bodor N, Swalm C, Farley T, Cohen D, et al. Neighborhood food environments and Body Mass Index: the importance of in-store contents. Am J Prev Med. 2009;37(3):214–9.

30. Zenk S, Powell L, Rimkus L, Isgor Z, Barker D, Ohri-Vachaspati P, et al. Relative and Absolute Availability of Healthier Food and Beverage Alternatives Across Communities in the United States. Am J Public Health. 2014;104(11):2170–8.

31. Lucan S. Concerning Limitations of Food-Environment Research: A Narrative Review and Commentary Framed around Obesity and Diet-Related Diseases in Youth. J Acad Nutr Diet. 2015;115(2):205–12.

32. Kelly B, Flood V, Yeatman H. Measuring local food environments: An overview of available methods and measures. Health Place. 2011 Nov;17(6):1284–93.

33. Cobb L, Appel L, Franco M, Jones-Smith J, Nur A, Anderson C. The relationship of the local food environment with obesity: A systematic review of methods. study quality and results. Obesity. 2015 Jul;23(7):1331–44.

34. Harris R, Minaker L, Mah C. The Nutrition Environment Measures Survey in rural communities: A systematic review. Unpubl Manuscr. 2017;

35. Accessibility-Remoteness (A-R) Index Summary Paper [Internet]. Newfoundland and Labrardor Statistics Agency; 2014. Available from: http://www.mae.gov.nl.ca/for/remoteness_index_information.pdf

84

36. North American Industry Classification System (NAICS) Canada [Internet]. Ottawa, Ontario: Statistics Agency of Canada; 2016. Available from: http://www.statcan.gc.ca/pub/12-501-x/12-501-x2016001-eng.pdf

37. Glanz K, Sallis J., Saelens BE, Frank L. Nutrition Environment Measures Survey in stores (NEMS-S): Development and evaluation. Am J Prev Med. 2007 Apr;34(4):282–9.

38. Honeycutt S, Davis E, Clawson M, Glanz K. Training for and dissemination of the Nutrition Environment Measures Surveys (NEMS). Prev Chronic Dis. 2010;7(6).

39. Lo B, Minaker L, Mah C, Cook B. Development and Testing of the Toronto Nutrition Environment Measures Survey-Store (ToNEMS-S). 2016;48(10):723–9.

40. Provincial and Territorial Guidance Document for the development of Nutrient Criteria for Foods and Beverages in Schools 2013 [Internet]. 2013. Available from: https://foodsecurecanada.org/sites/foodsecurecanada.org/files/pt_guidance_documenteng_-feb_18_2014.pdf

41. Gittelsohn J, Rowan M, Gadhoke P. Interventions in Small Food Stores to Change the Food Environment, Improve Diet, and Reduce Risk of Chronic Disease. Prev Chronic Dis PrevChronic Dis. 2012;9(Generic).

42. Calancie L, Leeman J, Pitts SJ, Khan L, Fleischhacker S, Evenson K. Nutrition-Related Policy and Environmental Strategies to Prevent Obesity in Rural Communities: A Systematic Review of the Literature, 2002–2013. Prev Chronic Dis. 2015;12(E57).

43. Minaker L, Mackenzie G, Nguyen N, Azagba S, Cook B, Mah C. An evaluation of the impact of a restrictive retail food environment intervention in a rural community pharmacy setting. BMC Public Health. 2016;16.

85

APPENDIX 6A Table 6A-1 NL Remoteness Index

Remoteness Classification

Access to Goods and Services

Highly Accessible Unrestricted Accessible Some Restriction Somewhat Accessible Considerable Restriction Moderately Remote Significant Restriction Remote Very Restricted Very Remote Little/ No Access

Table 6A-2 North American Industry Classifications for Store Types

North American Industry Classification

Definition

Supermarkets and other grocery (except convenience) stores

Establishments primarily engaged in retailing a general line of food such as canned, dry and frozen foods; fresh fruits and prepared meats, fish, poultry, dairy products, baked products and snack foods.

Convenience Stores Establishments primarily engaged in retailing a limited line of convenience items that generally includes milk, bread, soft drinks, snacks, tobacco products, newspapers and magazines

Gas stations with convenience stores

Establishments primarily engaged in retailing automotive fuels combined with the retail sale of a limited line of merchandise such as milk, bread, soft drinks and snacks in a convenience store setting.

86

APPENDIX 6B

Table 6B-1 Healthy Food Availability Score Item Availability Total Maximum

Score Fruits 1-3 varieties = 3pts

3-6 varieties= +2pts > 6 varieties = +3pts

6

Vegetables 1-3 varieties = 3pts 3-6 varieties = + 2pts > 6 varieties= + 3pts

6

Baked Goods Whole Wheat Bagel Or Whole Wheat English Muffin = 3pt

3

Bread Whole wheat bread available = 3pts

3

Fresh and Frozen Meats

Ground Beef= 2pts Chicken Leg= 2pts Pork Chops= 2pts Cod= 3pts

9

Meat Alternatives Canned Tuna= 3pts Yellow Split Peas= 3pts Kidney Beans= 3pts

9

Milk 2% /Skim =3pts 3 Eggs Eggs= 3pts 3 Cheese Cheddar Cheese =2pts

Deli Cheese = 2pts 4

Frozen fruit 1-2 varieties = 3pts 3-6 varieties = + 2pts >6 varieties= + 3pts

6

Frozen vegetables 1-2 varieties = 3pts 3-6 varieties = + 2pts >6 varieties= + 3pts

6

Frozen dinners Reduced-Fat=1pt Reduced Sodium=1pt

2

Canned Fruit 1-2 varieties = 3 pts 3-5 varieties = + 2pts >6 varieties= + 3pts > 1 kind packed in water without water = +1pt

7

Canned Vegetables 1-2 varieties = 3 pts 3-5 varieties = + 2pts >6 varieties= + 3pts> 1 kind without salt = +1 pt

7

87

Table 6B-1 Continued

Item Availability Total Maximum Score

Chips Baked = 1pt Lightly Salted= 1pt

2

Beverages Diet Pop = 1pt Reduced Sugar Sports Drink=1pt

2

Cereal Healthier cereal available = 3 pts > 1 kind healthier cereal = + 2 pts

5

Rice and Pasta Healthier rice available = 3 pts > 1 kind healthier rice = + 2pts Healthier pasta available = 3pts > 1 kind healthier pasta = + 2pts

10

Total: 93

88

Table 6B-2 Unhealthy Food Availability Score

Item

Availability Total Maximum Score

Baked Goods Muffin OR Tea Bun OR Apple Flip = 3pts

3

Bread White Bread = 2pts 2 Processed Meat Hot dogs = 3pts

Chicken Nuggets = 3pts Fish Nuggets = 3pts Bologna = 3pts Salt beef =3pts

15

Meat Alternatives Peanut Butter = 1pt 1 Milk Whole Milk = 1pt 1

Processed Cheese Cheese Slices = 2pts 2 Frozen Dinner Regular Frozen Dinner = 3pts 3 Chips Regular Chips= 3pts 3 Beverages Pop = 3pts

Sports Drink = 3pts 6

Cereals Sugar Cereal = 3pts > 1 variety sugar cereal = +2pts

5

Rice and Pasta White Rice = 2pts > 1 kind white rice = +1 pt White Pasta= 2pts > 1 kind white pasta = +1 pt

6

Total: 47

89

APPENDIX 6C

Table 6C-1 Healthy individual measure scores by store type

*Significant at p<0.01

Measure Mean Score (SD) Total Available

Points Supermarkets Convenience Stores

Fresh Fruit

4.7* (2.1) 2.4 (2.4) 6

Fresh Vegetable 5.6* (1.0) 4.4 (2.1) 6

Healthier Baked Goods

0.71 (1.3) 0.49 (0.38) 3

Whole Wheat Bread

2.65 (1.0) 2.16 (1.4) 3

Healthier Meat 6.4* (2.0) 3.8 (2.9) 9

Healthier Meat Alternatives

8.3 (1.7) 7.6 (2.5) 9

Low-Fat Milk 2.7 (1.0) 2.8 (0.8) 3

Eggs 2.8 (0.7) 2.5 (1.1) 3

Healthier Cheese 2.0 (1.0) 1.8 (1.2) 4

Canned Fruit 6.4* (1.3) 4.4 (2.4) 7

90

Table 6C-1 Continued

*Significant at p <0.01

Measure Mean Score (SD)

Total Available Points

Supermarkets Convenience Stores

Frozen Fruit 2.0* (2.2) 0.4 (1.1) 6

Frozen Vegetables 4.1*(2.4) 1.4 (2.0) 6

Healthier Frozen Dinner

0.8* (0.6) 0.3 (0.6) 2

Healthier Chips 1.6 (0.6) 1.6 (0.6) 2

Healthier Beverages

1.6 (0.5) 1.7 (0.5) 2

Low Sugar Cereal 4.1 (2.0) 3.5 (2.1) 5

Whole Wheat Rice

2.4* (2.4) 0.3(1.0) 5

Whole Wheat Pasta

2.5* (2.5) 0.3 (1.2) 5

91

Table 6C-2 Regular individual measures by store type

*Significant at p<0.0

Measure

Mean Score (SD)

Total Available Points

Supermarket

Convenience Store

Regular Frozen Dinner

1.9 (1.5) 1.5 (1.5) 3

Regular Chip Score

3.0 (0.0) 3.0 (0.0) 3

Regular Beverage

5.1 (2.1) 5.7 (1.1) 6

Regular Cereal 4.3* (1.7) 2.9 (2.3) 6

White Rice 2.6 (1.0) 2.1 (1.2) 5

White Pasta 2.8 (0.7) 2.8 (0.8) 5

Regular Baked Goods

2.7 (1.0) 2.2 (1.3) 3

White Bread 1.8 (0.7) 1.7 (0.7) 2

Processed Meat 12.5*(3.0) 8.4 (4.6) 15

Regular Meat Alt

0.9 (0.3) 0.7 (0.4) 1

Processed Cheese Score

1.9* (0.5) 1.4 (0.9) 2

92

CHAPTER 7: Improving Access to Nutritious Food in Rural Newfoundland and

Labrador: A Policy Analysis

93

AUTHORSHIP STATEMENT

TITLE: Improving Access to Nutritious Food in Rural Newfoundland and Labrador: A Policy Analysis AUTHORS: 1. Rebecca Harris (Lead Author), MSc Student, Community Health and Humanities, Memorial University of Newfoundland and Labrador 2. Dr. Catherine Mah, Associate Professor, School of Health Administration, Dalhousie University

Contributions

CM and RH proposed the concept for this manuscript. RH devised an implementation plan

and framework for the analysis. RH conducted background research, policy analysis and

conclusions. All drafts of the manuscript (including all sections) were written by RH. Initial

and intermediate versions were sent to CM for edits and suggested revisions.

94

7.1 ABSTRACT

Rural food environments have been characterized as disadvantaged because of the low

accessibility to full service grocers and a greater dependency on small stores that lack the

capacity to offer nutritious foods at affordable prices.

The objective of this analytic essay is to identify potential strategies to improve the

accessibility of nutritious food in Newfoundland and Labrador (NL). We proposed and

critically analyzed two fiscal policy options: 1) offering financial grants to support the

development of full service grocers in rural NL and 2) renovation grants and distribution

subsidies for small rural retailers to improve their capacity to stock healthy foods.

Both options address fundamental barriers to healthy food access in rural areas. However,

grants to increase the capacity of existing rural retailers to offer healthy items requires a

smaller upfront investment and more evidence is available on the success in influencing

the availability, purchase and consumption of nutritious foods. Implementation of this

option has increased feasibility because of a similar pilot program implemented by Food

First NL, Memorial University and Eastern Health. Multi-sectoral collaboration will be

essential to ensure the success and sustainability of the program.

95

7.2 INTRODCUTION AND BACKGROUND

With the overwhelming economic, morbidity and mortality related implications of the

rising rates of chronic disease, public health officials and policy makers are under added

pressure to reduce the “risk behaviors” contributing to non-communicable disease. Risk

behaviors include behaviors such as smoking, alcohol abuse, physical inactivity and

unhealthy eating, that increase the risk of developing chronic, non-communicable

diseases.1

Recent social and cultural shifts alongside evolution in the food industry have led to an

increase in the marketing, availability and affordability of highly processed, high fat, high

sugar foods and a growing dependence and accessibility to energy dense, nutrient poor,

items.2,3 These trends have created food environments that promote the consumption of

an unhealthy diet while making it increasingly difficult for consumers to make healthful

choices.2,3 The situation is exaggerated in rural communities where inadequate

accessibility to fresh and nutritious foods is compounded by geographic restrictions in

remote locations and a lack of public transportation, and increased food costs.4,5,6,7

Unlike alcohol and tobacco, the nutrients and energy from food and beverage are essential

for life, thus, regulating the consumption of food and beverage is more challenging than

for alcohol or tobacco.8,9 Food is also a significant component of social and cultural well-

being and it can be argued that executing authority over dietary behaviors imposes on

individual autonomy.8 Consequently, establishing and enforcing authoritative regulations

96

over individual consumption of food and beverage is complex and has been largely

disregarded.8 Opportunely, the use of fiscal policy tools including grants, subsidies and

tax, offer a proactive but malleable approach to modifying individual dietary behavior;

using the power of treasure10 to incentivize or discourage the consumption of certain

foods without infringing on public liberty.

The aim of this paper is to analyze fiscal policy as a population health intervention to

improve availability and affordability of nutritious food in Newfoundland and Labrador

(NL), where over 40% of the population live in rural or remote areas11 and over 3/4 of the

population is overweight or obese.12

We will begin by outlining the issue history of obesity and diet related disease and the

main barriers to creating healthy food environments in rural communities. Next, we will

provide a brief introduction into the role of fiscal policy in food environments, with an

overview of some existing fiscal strategies that have been implemented across other

jurisdictions. We will then review some of the action being taking at the federal and

provincial level to address to food access and healthy eating in NL. Finally, we will

propose and analyze two fiscal policy options aimed at improving equitable access to

healthy foods in rural NL, followed by our recommendations for future action.

97

Problem Statement

Obesity and the Food Environment

Obesity is a national and international public health crisis. Currently in Canada, the

number of people who are overweight or obese exceed the number who are of healthy

weights, with 61% of the national population classified as overweight or obese.12

According to the most recent data, NL has the highest prevalence of overweight and

obesity in the country where 77.5% of adults classified as overweight or obese.12

The prevalence of diet related diseases, including diabetes, cardiovascular disease,

cancer, osteoporosis and dental disease, are also approaching epidemic proportions.1

Newfoundland and Labrador has the highest prevalence of diabetes in Canada with 9.6%

of the provinces’ adult population facing type 1 or type 2 diabetes.13

Together, diet related diseases, including obesity, are the leading cause of death in

Canada and place an insurmountable economic burden on the health care system- costing

tens of billions of dollars each year.1  The estimate per capita total health care costs in

Newfoundland and Labrador was approximately $7256 in 2016.14 Although these

diseases are primarily preventable through a healthy lifestyle (including a nutritious diet)

a number of indicators suggest that diet quality is poor in NL. For example, only 30% of

Canadian adults and 18.5% of Newfoundland adults consume the recommended 5

servings of fruit and vegetables per day. 13

98

Understanding the complexity of factors influencing food choice is imperative to

developing successful policy intervention to modify dietary behavior and reduce the

burden of diet related disease. This includes recognition that to observe change in

individual behaviors there must be supportive environments in which to make healthy

choices.2,3,15 The food environment is a significant component of the physical

environment that influences diet. The food environment is the accessibility, availability

and affordability of food and beverage in a defined community or geographic area.16 It

can be investigated through the community food environment: the distance, proximity or

density of food retailers within a defined area (i.e. convenience stores, grocery stores or

fast food outlets) or through the consumer (retail) food environment: the availability,

affordability and quality of food available within food stores.15, 20 Although the field of

food environment research is relatively new, findings have consistently demonstrated that

those living more disadvantaged communities, such as low income neighbourhoods or

rural and remote communities are disproportionally exposed to unhealthy food

environments.4,5,6,7

Barriers to Building Healthy Food Environments in Rural Communities

Rural communities face a catalogue of health, demographic and socio-economic

disparities compared to those living in urban communities. Rural populations are aging,

they may have lower levels of educational attainment, they are more likely to be

unemployed, they may have higher prevalence of income from social assistance sources,

they may have more limited access to health care services, and have generally poorer

health status and higher mortality rates compared to their urban counterparts.17

99

Rural communities differ from even nearby urban areas. They have smaller populations

over a larger area of land, they are further from the metropolitan area and residents often

have to travel a significant distance for some of their basic necessities.5,17,18 Public

transportation is limited in rural communities therefore residents may have increased

reliance on nearby food retailers.17 As a result, many living in rural communities may be

forced to do the majority of their grocery shopping at smaller local stores, such as

convenience stores. This is particularly relevant to NL where there is the highest

proportion of convenience stores in rural areas (72%) of any province/territory in

Canada.19

Rural retailers face additional barriers to offering nutritious foods as a result of limited

opportunity to apply economies of scale. Many of the stores in rural communities are

small because of objective or perceived low consumer demand and lack of profitability

for a full-service grocery store.20,23 It is difficult for the small stores to stock healthier

options, such as fresh produce, because unlike larger stores they don’t have the benefit of

purchasing in bulk, they may face higher purchasing costs, and their rural location adds to

increased distribution expenses.21 Further, because of the small size and limited

infrastructure of corner stores, they often do not have the capacity to stock and display

fresh or frozen foods which can lead to produce and other perishable items deteriorating

quickly. 21,22

As a consequence of the economic and environmental barriers facing both retailers and

consumers, creating healthy retail food environments is a challenge in rural communities.

100

Access to healthy foods in rural NL is not an issue that is only relevant to a small portion

of the province; 41% of Newfoundland’s population live in rural communities.11

Implementing fiscal policies in the retail food environment is an area of growing interest

to population health decision makers. This paper examines whether fiscal policies can

relieve some of the specific barriers in rural areas and work to improve both availability

and accessibility of nutritious foods, toward improving diet related population health

outcomes.

7.3 FISCAL POLICY IN THE FOOD ENVIRONMENT

The relationship between diet and non-communicable diseases has been well established,

however, the ethical debate regarding individual or environmental factors’ relative

contribution to disease and the roles of government and industry in shaping these factors,

remains under dispute.2,23 Although some maintain that over indulgence and poor diet is a

personal choice, in which the individual acknowledges and accepts the potential

consequences of their behavior, others argue that the recent shifts in the food environment

including increased accessibility, availability and marketing of unhealthy food and

beverage has obstructed the ability for individuals to make rational and conscious

decisions regarding their diet.2,23 This idea is often framed under the concept of market

failure and used as a central argument in the advocacy for public health intervention in

the food environment.24,25 Market failure occurs when there is an inefficient distribution

101

of resources and the consumption of a good or services is not in the best interest of

society, thereby justifying government intervention.26

Fiscal policies have gained traction and support from decision makers in recent years

because they offer a balance between government stewardship and individual autonomy.

This concept which has been referred to as a “soft paternalism” approach by some

authors.25 That said, some fiscal policy can be perceived to be “harder” than others, most

notably taxation, while others, such as subsidization, can be perceived as more “soft” or

malleable approaches.25 Some of most common fiscal interventions that have been

proposed or implemented in the retail food environment are outlined below.

Unhealthy Food Tax

Taxation on alcohol and tobacco products has been well described as a successful public

health intervention 27,28 and has stimulated interest and advocacy for taxation on other

products, such as unhealthy food and beverages, in recent years. In the last decade, many

countries, including Denmark, France, Hungary, Mexico, among others, have begun

implementing taxes on sugar sweetened beverages and foods defined as “unhealthy” by

various standards.27,29,30,31 Evidence regarding the effectiveness of taxes have been mixed,

in part because the taxes accomplish different aims. Although many reviews on the topic

have concluded taxation is effective in producing modest changes in consumer behavior

and health outcomes, they have also demonstrated that price elasticity varies by item and

population, and is dependent on the type and amount of tax collected.27,28,30 Taxation is

among the few retail food environment interventions in the literature that target unhealthy

102

foods, however, a central argument against nutrition-oriented taxes is that they are

regressive and disproportionately burden low income individuals who spend a higher

proportion of their income on tax and are more likely to purchase taxed foods.9,27,28,30,31

More research is needed to establish alternative, evidence-based approaches to

intervening in consumer exposure to non-nutritious foods that may be more suitable for

rural communities and other populations with high proportion of low income residents.

Healthy Food Subsidies

Nutrition subsidies are based on the economic theory that by decreasing the cost of

targeted products relative to an alternative, consumers will be more likely to purchase the

targeted products. 9 That is, by decreasing the price of nutritious foods relative to the price

of energy dense foods high in fat, sugar and sodium, consumers will be more motivated to

make nutritious purchases.  Food subsidization has been shown to be a successful nutrition

intervention; a 2012 review of field experiments in healthy food subsidies reported that

all but one of the reviewed studies (19 of 20), found a significant increase in the purchase

or consumption of the subsidized foods.32 Similar to taxation, the success of the subsidy is

dependent on the level of discount applied, with more generous discounts demonstrating a

greater increase in purchasing and consumption of subsidized items.32,33 Although

effective, subsidies require substantial financial resources to maintain.

Development of Full Service Grocery Stores

Research has demonstrated a positive relationship between the proximity and density of

grocery stores and community diet and health patterns.4,34,35 The concept of building full

103

service grocery stores in disadvantaged communities is most often with the intent to

remove the geographic barrier to accessing quality, healthy foods and/or to decrease the

economic burden of purchasing these items in smaller, higher priced stores.20,36

Introducing a new retailer may also have the added benefit of creating local jobs while

increasing the economic vitality of the community.20,36 Government intervention in

bringing full service grocery stores to underserved communities can be operationalized in

many ways, including: documenting consumer demand and household spending power,

offering tax credits and regulatory incentives (expediting zoning or permit review) or

larger financial contributions in the form of non-repayable grants.20,36 A perceived

limitation of grocery store grants as a policy intervention is the significant up front and

long term investment which may pose more risk than other alternatives.

Healthy Corner Store Programs

Because of a perception of low consumer demand, the high cost of distribution, small size

and limited capacity of small rural stores, offering fresh healthy foods can be

challenging.21 In efforts to help retailers overcome these barriers and increase access to

healthy foods in underserved areas, the concept of the “healthy corner store” emerged.

Healthy corner store projects are now widespread across the United States and are

becoming more prevalent in Canada.37 Healthy corner store projects are most often

supported through government grants and can be small interventions such as

merchandising and marketing campaigns to promote healthy eating or larger

infrastructural changes and equipment installation that increase a stores’ capacity to stock

healthy foods.37,38 Although evaluations of urban healthy corner store programs have been

104

mostly optimistic, they have been primarily designed for implementation in urban centers

and most do not address some of the key barriers to food access in rural communities,

such as costly distribution. 22,39

7.4 LOCAL POLICY

Improving population health outcomes was named as one of the top three priorities in the

2014-2017 Newfoundland and Labrador provincial strategic wellness strategy.40

Unfortunately, there are currently no direct provincial programs or policies in place aimed

at improving equitable access to healthy foods in rural NL. There is however, some direct

and indirect action being taking at the federal and provincial level to address to food

access and healthy eating. We will review some of these strategies below.

Provincial Action

Provincial Healthy Eating Framework

In response to rising rates of obesity and diabetes, in 2006 the province published “Eating

Healthier in Newfoundland and Labrador” a provincial food and nutrition and framework

action plan.41 The action plan was designed to guide government, organizations and other

key stakeholders in decisions regarding healthy eating and was intended to be

implemented in phases. Phase one was published in 2006 and described goals and

strategies that would take place from 2005-2008. The framework vision reads “All

residents of Newfoundland and Labrador will have reasonable access to an adequate,

nutritious and safe food supply and a supportive, comprehensive network of food and

105

nutrition services.”41 The goals of the framework include contributing to individual and

community well-being by supporting health promotion strategies directed toward healthy

eating and physical activity, supporting measures that will allow access to adequate food

supply and nutrition services to vulnerable populations, supporting the development,

production, marketing and distribution of healthy foods and supporting food and nutrition

research.41

The development of a framework involves significant investment of time and resources.

The framework development involved a multi-sectorial approach and consultation was

made with over 140 groups and over 800 recommendations and comments were given.41

Unfortunately, approximately a decade later, phase one of the framework has been the

only one published. Although some of the actions have been implemented, such as the NL

school nutrition policy, the majority of the actions, including collaborating with food

producers and suppliers on initiatives that support the availability of healthy foods, have

not moved forward.

Community Healthy Living Fund

In 2015, the provincial Department of Health and Community Services introduced the

community healthy living fund.42 The community healthy living fund replaced four

previous existing grant programs including the senior’s recreation grant, provincial health

and wellness grant, and the community recreation support program. The community

healthy living fund is divided into three categories: capacity building, programs, and

supportive environments.

106

Application for participation in the grant program is competitive but if chosen, applicants

can receive up to 20,000 in one-time funding. The 2017 provincial budget has allocated

1.79 million dollars to support initiatives under the community healthy living fund but it

is unclear which areas of the program will be given the most priority.42 In 2016, the

funding was distributed across almost 400 organizations and many of the larger grants

were allocated to sports and recreation.43

Because the scope of the grant program is so broad, it may lead to smaller investments

made in a wide range of areas, instead of larger, more impactful change in high priority

areas. Unfortunately, minimal to no evaluation appears to have been done to determine

the success of the program and it is difficult to determine if investments are being made

into sustainable interventions to improve long term health outcomes of the province.

Federal Action

Nutrition North

In 2011, in response to substantial and increasing food insecurity rates in Northern

Canada, the federal government announced Nutrition North Canada (NNC).44 NCC is a

Health Canada program that provides subsidies to retailers and suppliers to relieve the

high cost of stocking and distributing perishable foods to the North. Retailers must apply

to participate and are selected based on eligibility criteria and need; they are then

responsible for forwarding the full subsidy on to the consumers. The subsidy is based on

current food cost and distribution expenses and only applies to an approved list of

foods.44

107

Nutrition North Canada is a significant national investment. The government invests

approximately 60 million dollars each year to NNC.44 The government has promoted the

success of the program, stating that since 2011, the cost of food for a family of four has

decreased approximately 5% per month despite the 25% increased cost of shipping

eligible items to Northern communities.44

Despite communication about the program’s success and the significant financial

contribution made toward the program, NCC has been criticised.45 A recent Auditor

General Report has identified a lack of transparency and retailer accountability as

significant issues.45 In efforts to increase accountability and transparency, the government

announced a point of sales system that will show consumers where and when the NCC

subsidy has been applied; all NCC retailers were mandated to implement the sales system

by 2016.44

Unfortunately, according to the program’s eligibility criteria, only those communities

who have used Food Mail and lack year-round surface transportation can participate.44

Based on these criteria only a limited number of NL communities are eligible, leaving

many rural communities unassisted.

The Way Forward

Public health research has produced evidence of significant disparity between urban and

rural food environments.4,5,6,7 It has also been suggested that limited access and

availability to healthy nutritious food is contributing to poor diet quality and the

108

development of overweight and obesity.5,18,22,46,47,48 This is of concern for NL where a

significant proportion of the population live in rural communities.11 This concern, in

combination with the alarming rates of overweight and obesity and other diet related

disease in NL, demonstrate an urgency for action.

The current strategies in place to promote healthy eating and improve the health and well-

being of the province are a step in the right direction but have failed to make a significant

impact on provincial diet related health outcomes. The province continues to have some

of the leading rates of overweight and obesity, diabetes and hypertension in Canada13

which indicates that what is being done, is ineffective or insufficient. The provincial

strategies currently in place have a broad scope and as a result, it is difficult to make an

influential impact on multiple target areas. A more direct and prioritized approach toward

reducing barriers and challenges to healthy eating in rural NL is needed.

7.5 POLICY ALTERNATIVES

The policy options proposed below present potential fiscal strategies to improve the retail

environment in efforts to improve access to healthy affordable food and make healthy

choices easier for those living in rural communities across NL.

109

Option 1: Rural Food Retailer Development Grant

Developing and maintaining a business in any area is associated with significant upfront

investments and high operating costs. Due to the low-density populations in rural

communities, there is often a perception of low profitability for businesses owners. As a

result, the majority of rural communities in NL lack a full-service grocery store.49

However, unlike small corner stores, full service grocery stores are larger and have

greater structural capacity to offer a wider variety of foods, they are able to make larger

purchases therefore face lower food costs and often have higher revenue to cover the

increased cost of distribution.20,36,50 By offering a grant for the development of full

service grocery stores in rural communities there is potential to attract business owners to

expand in underserved rural areas and improve the access to healthy nutritious food.20,36,50

This proposed program is modeled on the Healthy Food Financing Initiative (HFFI)51, a

retail food environment program implemented in the United States, and is described

below incorporating adaptations that will increase suitability for implementation in rural

NL.

In this program, prospective food retailers could apply for the grant and would be selected

based on priority eligibility criteria, including their ability to demonstrate concern for the

health of the community, motivation to improve access to healthy foods, while also

presenting a sound and feasible business plan. Because the size of rural communities are

small, this program would target small full service grocers which may include

cooperative, independent or chain grocers. In order to qualify as a full-service grocer,

retailers would need to fall within the definition of supermarket and grocery store as

110

defined by North American Industry Classification System.52 This would include offering

fresh, frozen and canned foods including fruits, vegetables, meat, poultry, dairy and

baked products. In order to be feasible, it would be reasonable for this program to target

fewer, higher populated rural communities of 800-1000 residents.

The proposed development grants would be a one-time contribution and would cover up

to 20% of development costs. Distributers would be required to sell foods at fair prices,

stock a minimum proportion of healthy foods and display healthy foods visibly and

attractively. The retailers would be monitored and evaluated on their compliance during

annual food safety inspections and would be subject to fines for non-compliance.

Option 2: Healthy Corner Store Program: Small Food Retailer Renovation Grants and

Distribution Subsidies

Because of the high cost of distribution, the size and limited capacity of small rural stores,

offering fresh healthy foods can be challenging.21,22,45 A program that offers store

renovation grants and food distribution subsidies could help rural food retailers overcome

these barriers by decreasing financial risk for rural food retailers and increase their ability

to offer healthy, nutritious food to their community. The following proposed program is

modeled after existing healthy corner store programs3853 that have been implemented

across North America, and is described below incorporating adaptations and additions

that will increase suitability for implementation in rural NL.

111

In this program, rural retailers would apply to participate based on specific eligibility and

could apply for either the food distribution subsidy, the renovation grant, or both. The

food distribution subsidy would be given to food retailers to compensate for the high cost

of distribution and the retailers would be required to hand the subsidy down to consumers

by offering a price that is equivalent to those in larger metropolitan areas. The distribution

subsidy would be applied to a list of approved healthy foods, for example, those that are

part of the pre-established Newfoundland and Labrador Nutritious Food Basket.54 Items

included in the Nutritious Food Basket are based on Canada’s Food Guide

recommendations, it also considers provincial purchasing patterns to ensure that the list

includes culturally appropriate foods. The subsidy would be calculated using a formula

including food weight and current food and distribution costs.

If a retailer wishes to receive a renovation grant they should first participate in the

subsidy program for six months in order to demonstrate they would make a good

candidate for renovation. The renovation grants would then be given to retailers who

require structural or equipment changes to increase their capacity to offer a wider variety

of healthy foods. Because public health food inspectors are an existing resource with

experience evaluating the retail food environment, they could be a valuable government

asset as a practitioner to assess the store to determine the necessary improvements.

Acceptance into the program would be competitive and based on eligibility criteria and

willingness to participate. In order to be eligible for participation in the program, for

example, it would be reasonable for the retailer to be located preferably in a somewhat

112

accessible, moderately remote, remote, or very remote area, as defined by the NL

Remoteness Index55 and to hold and existing NL food establishment license. As part of

their application, retailer owners could be requested to state their motivations for

participating in the program, including: describing their barriers to offering fresh

perishable items, demonstrating a sense of concern for the lack of healthy options in their

community, and recognizing their role in making a positive impact on the health and

well-being of their community.

A set of accompanying program regulations could include: forwarding the full

distribution subsidy to the consumer, stocking a minimum proportion of healthy options,

displaying healthy foods in a visible and attractive manner, display program marketing

materials and monitor the sales of subsidized healthy foods. In this instance, it would be

important for retailers to participate in enforcement strategies to monitor and evaluate

compliance. Monitoring could be conducted by public health food inspectors who

integrate such inspections as part of routine food safety inspections. Retailers would also

be required to prepare self-reports, i.e., to submit their sales of subsidized foods and

demonstrate that the subsidies have been passed on to consumers. Consequences of non-

compliance could include fines, and a dismissal in program participation.

113

7.6 POLICY ANALYSIS

Analytical Framework

In the following section, we will analyze the policy options in further detail,

supplemented by a critical examination of relevant scientific and grey literature. We will

use the framework shown in Figure 7-1, from the National Collaborating Centre for

Healthy Public Policy (NCCHPP)52, to analyze the effects and implementation of the

proposed programs. NCCHPP’s policy analysis framework integrates the decision-

making needs from both the public health perspective (effectiveness, unintended effects

and equity) and policy maker perspectives (cost, feasibility and acceptability). It offers

guidelines within a flexible framework that is adaptable for varying degrees of time

resources, making it a logical and suitable framework for this context.52 A summary of

policy option considerations can be found at the end of the analysis, in Table 7-1.

Figure 7-1 The Framework for Analysis of Proposed Policies

114

Effectiveness Rural Food Retailer Development Grant

Preliminary research has established a positive relationship between proximity and

density of grocery stores with consumer behavior, diet and health outcomes.4,34,35 In a

multi-state U.S study of more than 10,000 adults, adults living in neighborhoods with

supermarkets or grocery stores had lowest rates of overweight (60–62%) and obesity

(21%) while those living in neighborhoods with no supermarkets and access to only

convenience stores and/or smaller grocery stores, had the highest rates (73–78%

overweight and 32–40% obesity).34 These findings provide evidence that supermarkets

and access to nutritious food plays a significant role in building healthier food

environments.

This evidence has gained the attention of policy makers and decision makers in the U.S

and has led to the development of Healthy Food Financing Initiative (HFFI), a program

designed to bring grocery stores and other healthy food retailers to underserved

communities across the United States. Unfortunately, although HFFI has developed or

supported over 1000 grocers since the implementation, there has been little to no

evaluation to demonstrate the effectiveness of these efforts.57

The few evaluations that have been done were conducted in urban centres and results

contradict expected outcomes. In two separate evaluations by each Cummins et al.57 and

Dubowitz et al.58, the introduction of a HFFI supermarket did not appear to improve

115

purchasing patterns, dietary consumption or body mass index. A third evaluation, by

Ghosh-Dastidar and colleagues59, evaluated overall community food availability

following the introduction of a grocery store in an underserved community; results

demonstrated that there was only a moderate and insignificant increase in community

availability of fruits, vegetables and other healthy items and it was paired with a moderate

increase in unhealthy food.59

Unfortunately, evaluations of grocery store development in rural communities are lacking

and it is difficult to know if the results from urban centers are generalizable to rural

communities. Many communities in NL do not have a grocery store or supermarket49

and over 70% of stores in rural communities are convenience stores or corner stores.19

For this reason, we predict this program could potentially have a more powerful effect in

rural NL communities but gaps in knowledge surrounding the success for these program

in rural and urban areas remain.

Small Food Retailer Renovation Grants and Distribution Subsidies

Because healthy corner stores have become such a prominent food environment

intervention, there have been numerous evaluations regarding the effectiveness of this

approach in improving food availability, consumer purchasing, and diet.22,39 Despite the

growing number of evaluations, those in rural communities are lacking.22, 39 Reassuringly,

evaluations conducted in urban communities have illustrated positive outcomes. A review

by Gittelsohn identified 16 articles and grey literature evaluating small store interventions

between 2005 and 2010.22 The majority of interventions reported an increase in

116

availability and sale of promoted foods with some trials observing 25 to 50% increase in

produce sales.22 Increases in consumer food and health related knowledge, as well as an

increased perceived accessibility and intention to buy healthy foods, were also

observed.22,25

Subsidies have also been shown to be a successful nutrition intervention; a 2012 review

of field experiments in healthy food subsidies reported that all but one of the reviewed

studies (19 of 20), found a significant increase in the purchase or consumption of the

subsidized foods.32 The success of the subsidy is dependent on the level of discount

applied, with more generous discounts demonstrating a greater increase in purchasing and

consumption of subsidized items.32,33 In school settings, discounts of 50% have

demonstrated up to 93% increase in sale of targeted items.30

Summary of Considerations

Although research has shown a positive relationship between proximity and density of

grocery stores with many outcomes of interest (consumer behavior, diet, health

outcomes), there is a lack of evidence that indicates introducing a new grocery store to an

underserved area will lead to an improvement in these areas. More evidence is available

for the effectiveness of healthy corner store programs and healthy food subsidization in

improving availability, purchasing, consumption and perceptions of healthy foods.

117

Unintended Effects

Rural Food Retailer Development Grant

One of the primary concerns surrounding the development of larger grocery stores in

rural or urban food deserts is the potential increase in the availability of unhealthy foods.

Although supermarkets have been considered generally “healthful” food retailers, our

previous research in NL, as well as similar research in the field, has shown that stores

with higher availability of healthy foods also tend to offer a higher exposure to unhealthy

foods.60,61 Consequently, increasing accessibility to unhealthy foods could lead to an

unintended increase in the purchasing and consumption of these items. However,

accompanying regulations to sell foods at fair prices, stock a minimum proportion of

healthy foods, and display healthy foods attractively, could help mitigate the potential

effects of an increase in the availability of unhealthy items.

Small Food Retailer Renovation Grants and Distribution Subsidies

Although healthy corner stores do not increase the availability of unhealthy food, there is

a chance that consumers may simply add subsidized foods to their diet rather than

substituting them in place of less healthful foods.28 For instance, an intervention based

study conducted in Australia placed a 20% subsidy on fruits, vegetables, bottled water

and diet drinks and while they observed a 20% increase in the purchase of fruits and

vegetables, they also observed a 13% increase in the purchase of unhealthy foods.62 With

that said, it is unclear if this would occur in rural communities where the price elasticity

(a measure of consumer response due to change in price) may differ based on overall

higher food prices. If the price of unhealthy food items remained unchanged, is possible

118

that rural consumers would take the opportunity to reduce food spending and therefore

substitute (rather than supplement) unhealthy items with subsidized healthy items.

Summary of Considerations

Because the healthy corner store program does not increase exposure to unhealthy foods

and because of the high baseline food prices and reduced purchasing power in rural

communities, we feel that the increase in unhealthy food purchasing is less likely to occur

in this program compared to the retailer development grant.

Equity

Equity is one of the strongest values when implementing a community based policy and it

lies at the core of NL’s food access issue. As provincial taxpayers, rural and urban

residents both have an interest in provincially implemented programs.

Both proposed programs are designed to prioritize rural communities and it is possible

that urban residents may be less supportive and view either program as inequitable.

However, rural communities face numerous health, demographic, and socio-economic

disparities compared to their urban counterparts and rural retailers also face barriers and

disadvantages to building and maintaining a successful business that stores in urban

centres may not. For this reason, we feel rural prioritization in both proposed programs is

justified.

119

Summary of Considerations

Both programs equally prioritize investment in rural populations. Within rural

communities, participation in the programs are competitive and not universal. Both

programs take similar steps to ensure the most equitable distribution of the resources by

awarding grants based on eligibility criteria, willingness to participate, feasibility and

need.

Cost

It was estimated that in 2017 the government would spend approximately $7377 per

capita on health care costs in Newfoundland and Labrador.14 The latest estimate of

obesity related health care costs is from 2006 when annual cost of overweight and obesity

was $126 million dollars in Newfoundland and Labrador.41 The prevalence of overweight

and obesity has risen steadily since then, therefore, it is expected that the related costs

have followed a similar trend. These figures demonstrate that investing in effective and

sustainable provincial public health strategies to prevent and reduce overweight and

obesity is necessary and justifiable.

Rural Food Retailer Development Grant

Implementing a small full service grocer can be costly and have been estimated to cost up

to 1 million dollars for capital costs including, land, building, amenities and equipment.63

Based on the commitment to cover 20% of development costs, government contribution

would be approximately $200,000 per program. This estimate is consistent with

contributions made in the Healthy Food Financing Initiative (HFFI) where grants range

120

from $170,000 to $800,000 depending on the size of the store and community.64 Based on

the estimate of $200,000 contribution from government, the development of 10 small full

service grocers would be $2 million dollars. Retailer development grants require larger

upfront investments but would only require a one-time contribution and will serve

multiple communities.

Small Food Retailer Renovation Grants and Distribution Subsidies

Reports from similar infrastructure corner store renovations have documented an average

cost of approximately $1400 depending on the type of renovations or equipment

required.65 The majority of the costs associated with this program would be dedicated to

supporting the food subsidy program. These costs will be based on current food and

distribution costs so they would vary. In the federal food subsidy program (NCC), cost

were documented at approximately $800 per person annually.44 Based on this estimation,

providing subsidization for a community of 300 people would cost roughly $240,000

annually and about $5 million annually to support 20 communities of this size. However,

it is important to note that NCC serves very remote northern communities where air and

boat distribution is necessary, therefore, we can predict a lower cost for a provincial

program serving communities accessibly by land. Further, research on cost effectiveness

on similar healthy food subsidy programs in rural areas have been determined to be cost

effective based on the annual costs disability adjusted life years saved.66

121

Summary of Considerations

While the investment in either program could be justified, the large up-front investment in

grocery store development poses considerably more risk. The healthy corner store

program has smaller upfront investments and the investment in subsidies could be

adapted depending on program success.

Feasibility

The policy options proposed here would fall generally under the mandate of the

Department of Health and Community Services but would require a multi-sectoral

approach. Improving access to healthy affordable food and increasing the economic

vitality of rural communities aligns with the priorities of several government departments

and many provincial organizations including the Department of Tourism, Culture,

Industry and Innovation (TCII), Municipalities NL, Food First Newfoundland and

Memorial University. Partnering with these organizations will help ensure the programs

are feasible and sustainable.

Rural Food Retailer Development Grant

Building rural food retailers would require more upfront investments in time and financial

resources. This program also has a significant business and economic component and

would require partnering more closely with TCII, particularly the sector of Regional

Economic Development.

122

We would seek to partner with Regional Economic Development for support during

outreach to potential retailers and evaluating the strength of retailer applications and

business proposals. They could also provide guidance and support regarding site

allocation, land and building permits, and provision of funding.

The Department of Health and Community Services would be primarily responsible for

developing nutrition parameters surrounding stocking, displaying and pricing. These

guidelines could be overseen by a provincial Registered Dietitian and informed by similar

strategies implemented in the Healthy Corner Store NL pilot project19 (outlined in more

detail below). Scientific literature and the numerous toolkits, handbooks and

implementation guides developed as part of the Healthy Food Financing Initiative, could

also be consulted during guideline development.36 Once implemented, monitoring and

evaluation of the required guidelines could be facilitated by Service NL and food

establishment inspectors.

Economic support for this program could be provided by The Department of Health and

Community Services but could be supplemented by related grants through TCII under the

Regional Development Funding.67

Small Food Retailer Renovation Grants and Distribution Subsidies

One of the benefits of the healthy corner store program is that it can work with existing

retailers, infrastructure, mechanisms, and partnerships. Key partners in the planning and

123

implementation of this program would be the Food Policy Lab, Eastern Health and Food

First NL.

In 2015, The Food Policy Lab at Memorial University received federal funding for the

development of a pilot healthy corner store in rural NL.68 During this initiative, Memorial

University partnered with Eastern Health NL, the largest health authority in NL, and Food

First NL, a provincial non-profit organization whose motivation is to improve food

insecurity through promoting community based solutions to improve access to healthy

food across the province. The pilot program was implemented in 2015 and can provide

the proposed program with a healthy corner store model and offers a valuable opportunity

to expand on what has been already done including community outreach, networking and

increased awareness of the importance of food retailers in creating healthy communities.

Evaluations of the pilot will help to identify potential barriers and inform improvements

to future program implementation. This partnership may also provide an opportunity to

conduct program evaluation in an academic capacity, which will enhance the rigour of

program evaluations and provide results and that can be disseminated to knowledge users

more widely.

The pilot healthy corner store does not have an integrated distribution subsidy program,

therefore, support and guidance for this element of the policy could be sought from

federal sources involved in Nutrition North Canada as well as from provinces who have

implemented similar programs and can provide feedback on key learnings.69 Working

with a provincial registered dietitian to create program guidelines surrounding approved

124

foods and the advertising, marketing and display of healthy foods, would also help ensure

the program is founded on evidence-based practice.

Financial support for the healthy corner store program could come from the Department

of Health and Community Services with support for store assessment and adherence

evaluation coming from Service NL food establishment inspectors.

Summary of Considerations

Multi-sectoral partnership is essential for success in both proposed programs. Because the

healthy corner store program works with existing retailers and infrastructure and could

follow in the steps and key learnings from the pilot healthy corner store, this program

offers a more feasible implementation.

Acceptability

The proposed policy options use fiscal instruments to create a health promoting

environment and encourage consumers to make healthy choices. In contrast to policy

instruments that use central authority, these options allow consumers and retailers to

maintain their autonomy when making food choices. However, both programs would

require a significant investment of public resources, consequently, may represent trade-

offs between financial support for new or existing health and wellness programs. Such

redistributive considerations have potential political implications as well as affecting

stakeholder support for those in decision making roles. This concern is valid but should

be offset by support from other key stakeholders in both the public and private sectors

125

who will recognize the value of a program to improve the both the built environment and

the health and wellbeing of the province (a win-win approach). The following section

analyzes some of the key interests at play.

Department of Tourism, Culture, Industry and Innovation (TCII) and Municipalities NL

The Department of Tourism, Culture, Industry and Innovation (TCII) is the department

responsible for supporting the development of economically sustainable communities36 and

Municipalities NL is an independent advocacy organization for municipalities across the

province.37 These organizations will be key partners in assisting retailers in preparing

program applications, developing business plans and other skills essential for implementing

an economically successful project.37 Both organizations have made revitalizing small rural

communities a priority for their organization and improving the rural retail environment

can help them achieve this. 67,70

Food First Newfoundland and Eastern Health NL

Food First Newfoundland is a provincial non-profit organization whose mission is to

improve food insecurity through promoting community based solutions to ensure access

to adequate healthy food across the province. Food First has contributed to many

initiatives to improve food access across the province and is one of the key partners in the

healthy corner store pilot project.68 Eastern Health NL is the largest regional health

authority in the province, serving over 300,000 individuals across 21,000 km2.71 Eastern

Health employees over 50 registered dietitians who work in a variety of contexts and

work collaboratively with other health professionals and communities using strategies

126

that promote healthy eating and help prevent diet- related diseases.72 Eastern Health is

also partnering with Food First and Memorial University on the Healthy Corner Store NL

project. Both Food First and Eastern Health are already invested in improving food access

through the retail food environment and their support will be essential in implementing a

successful program at the provincial level.

Rural Food Retailers & Community Members

Depending on the program implemented, acceptance from existing rural retailers will

vary. Full service grocers could present a threat to small rural retailers and could

potentially cause some stores to close. For example, in an evaluation by Ghosh-Dastidar,

the introduction of a full service grocery store actually reduced the availability of healthy

foods in surrounding stores, who could not compete with low prices offered in the larger

store.59 Larger grocers could also be seen an effort to urbanize rural communities. This

may lead to some opposition from rural business owners and supporting community

members.

Summary of Considerations

The objectives of both proposed are similar and therefore could expect to be accepted

equally by most stakeholders. However, there are some reason for concern regarding

reactions from local retailers and some community members on the introduction of a full-

service grocery in the community. The healthy corner store program offers a less invasive

intervention and works with existing, local retailers and enables rural dwellers to support

their community members.

127

Table 7-1 Summary of Policy Considerations

Rural Food Retailer Development Grant

Small Food Retailer

Renovation Grants and Distribution Subsidies

Eff

ects

Effectiveness

-Lack of evaluation in rural areas -Evaluations in urban communities were only moderately effective

-Lack of evaluation in rural areas -Evaluations in urban areas have demonstrated the effectiveness in improving food availability, consumer purchasing, and diet

Unintended Effects

-Potential increase in the availability of unhealthy food

- Consumers may add subsidized foods to their diet rather than substituting them in place of less healthful foods

Equity -Prioritizes rural

communities -Prioritizes rural communities

Impl

emen

tatio

n

Cost

-Significant upfront cost but no cost to maintain -Lack of evidence for success in rural communities, therefore high risk investment

-Smaller upfront cost but subsidy will require continuous financial support - Evidence of program success and some evidence of cost effectiveness, therefore less risk

Feasibility

-Requires more upfront investments in time and financial resources, new partnerships

-Works with existing retailers, infrastructure, mechanisms, and partnerships

Acceptability

-Could pose a threat to small stores, may lead to some opposition from rural business owners

-Less invasive intervention and works with existing, local retailers and enables residents to support community members

128

7.7 SUMMARY AND RECOMMENDATIONS FOR FUTURE ACTION

It is evident that a solution to equitable food access in rural areas is imperative to

improving diet related health outcomes. The policy options proposed present strategies to

improve the retail environment in efforts to improve access to healthy affordable food and

make healthy choices easier for those living in rural communities. Based on the factors

and trade-offs presented in the analysis above, we would argue that the more suitable

policy option for NL among the fiscal interventions examined would be the small food

retailer renovation grant and distribution subsidy program.

Grants designed to offer incentives for the development of larger, full service grocery

stores in underserved communities requires significant upfront investment and evidence

on their success is lacking. There is also a chance this approach may not be accepted by

rural community members and could unintentionally increase accessibility to unhealthy

food.

Alternatively, administering grants to increase the capacity of existing rural retailers to

offer healthy items requires a smaller upfront investment and more evidence is available

on the success of healthy corner store interventions. This strategy is also predicted to be

more accepted by rural dwellers as it builds on existing infrastructure to improve

accessibility, while contributing to the economic vitality of the community. The Healthy

Corner Store NL pilot can help to improve feasibility and guide implementation of the

program.

129

Nutrition subsidies designed to mitigate the increased cost of nutritious food due to

distribution and purchasing expenses address significant barriers to food access in rural

communities and have proven to be successful in influencing the purchase and

consumption of nutritious foods. Although there are several known challenges with

Nutrition North Canada, its implementation offers valuable lessons and can be used as a

learning platform for future, smaller scale, more inclusive distribution subsidy programs.

Conclusion

The current strategies in place to promote healthy eating and improve the health and well-

being of Canadians have failed to make a significant impact on obesity and diet related

health outcomes in NL. The prevalence of diet related chronic disease continue to rise,

indicating that what is being done is ineffective or insufficient, calling for more direct and

prioritized action. It is clear that reducing barriers and challenges to healthy eating in

rural communities is a crucial step, and fiscal policy levers offers an influential and

potentially autonomy-preserving approach to achieve this.

130

REFERENCES

1. Public Health Agency of Canada. How Healthy Are Canadians? A Trend Analysis of the Health of Canadians from a Healthy LIving and Chronic DIsease Perspective [Internet]. 2016 Dec. Available from: https://www.canada.ca/content/dam/phac-aspc/documents/services/publications/healthy-living/how-healthy-canadians/pub1-eng.pdf

2. Roberto C, Swinburn B, Hawkes C, Huang T, Costa S. Patchy progress on obesity prevention: emerging examples, entrenched barriers, and new thinking. Lancet. 2015 Feb;385(9985):2400–9.

3. Story M, Kaphingst K, Robinson-O’Brien R, Glanz K. Creating healthy food and eating environments: Policy and environmental approaches. AnnuRevPublic Health. 2008;29:253–72.

4. Larson N, Story M, Nelson M. Neighborhood environments: Disparities in acess to healthy foods in the U.S. Am J Prev Med. 2008 Jan;36(1).

5. Liese A, Weis K, Pluto D, Smith E, Lawson A. Food store types, availability and costs of food in a rural environment. J Am Diet Assoc. 2007 Nov;107(11):1916–23.

6. Lendarson J, Hansen A, Hartley D. Rural and Remote Food Environments and Obesity. Curr Obes Rep. 2015 Mar;4(1):46–53.

7. Jithitikulchai T, Dean W, Sharkey J. Variations in the availability and price of healthier food options by store type and urban-rural setting. 2012 Dec;7(4):381–400.

8. Mello MM, Studdert DM, Brennan TA. Obesity — The New Frontier of Public Health Law. N Engl J Med. 2006 Jun 15;354(24):2601–10.

9. Powell L. Food prices and obesity: evidence and policy implications for taxes and subsidies. Milkbank Q. 2009 Mar;87(1):229–57.

10. Hood C. The Tools of Government. Chatham House Publishers; 1986.

11. Statistics Canada. Population, urban and rural by province and territory (Newfoundland and Labrador) [Internet]. 2012. Available from: https://www12.statcan.gc.ca/census-recensement/2011/ref/dict/geo042-eng.cfm

131

12. Statistics Canada. Table 105-2023 - Measured adult body mass index (BMI) (World Health Organization classification), by age group and sex, Canada and provinces, Canadian Community Health Survey - Nutrition, occasional, CANSIM [Internet]. [cited 2018 Jan 26]. Available from: http://www5.statcan.gc.ca/cansim/a47

13. Statistics Canada. Statistics Canada. Table 105-0508 - Canadian health characteristics, annual estimates, by age group and sex, Canada (excluding territories) and provinces, occasional CANSIM [Internet]. 2017. Available from: http://www5.statcan.gc.ca/cansim/a47#F14

14. Canadian Institute of Health Information. National Health Expenditure Trents, 1975 to 2016 [Internet]. 2016. Available from: https://secure.cihi.ca/free_products/NHEX-Trends-Narrative-Report_2016_EN.pdf

15. Glanz K, Sallis J, Saelens B, Frank L. Healthy nutrition environments: Concepts and measures. Am J Health Promot. 2005 Jun;19(5):330–3.

16. Minaker L, Shuh A, Olstad D, Engler-Stringer R, Black J, Mah C. Retail food environments in Canada: A scoping review. Can J Public Health. 2016 Jun;107(1):5344.

17. Crosby R, Wendel M, Vanderpool R, Casey B. Rural populations and health determinants, disparities, and solutions. 1st ed. Wiley; 2012.

18. Michimi A, Wimberly M. Associations of supermarket accessibility with obesity and fruit and vegetable consumption in the conterminous United States. Int J Health Geogr. 2010 Oct;9(49).

19. Memorial University. Healthy corner stores NL [Internet]. 2015. Available from: http://www.yaffle.ca/a/viewProject/id/1936

20. Getting to Grocery: Tools for Attracting Healthy Food Retail to Underserved Neighborhoods [Internet]. ChangeLab Solutions; 2012 [cited 2017 Nov 25]. Available from: http://www.changelabsolutions.org/sites/default/files/documents/Getting_to_Grocery_FINAL_20120514.pdf

21. Providing Fresh Produce in Small Food Stores: Distribution Challenges & Solutions for Healthy Food Retail [Internet]. ChangeLab Solutions; 2014. Available from: http://www.changelabsolutions.org/sites/default/files/Fresh-Produce-Distribution-Small-Food-Stores_FINAL_20140131.pdf

22. Gittelsohn J, Rowan M, Gadhoke P. Interventions in Small Food Stores to Change the Food Environment, Improve Diet, and Reduce Risk of Chronic Disease. Prev Chronic Dis PrevChronic Dis. 2012;9(Generic).

132

23. Brownwell K, Kersh R, Ludwig D, Post R, Puhl R, Schwartz M. Personal Responsibility and Obesity: A Constructive Approach To A Controversial Issue. Health Affairs. 2010 Mar;29(3).

24. Karnani A, Mcferran B, Mukhopadhyay A. The Obesity Crisis as Market Failure: An Analysis of Systemic Causes and Corrective Mechanisms. J Assoc Consum Res. 2016;1(3):445–70.

25. Swinburn B. Obesity prevention: the role of policies, laws and regulations. Aust NZ J Public Health. 2008 Jun;5(12).

26. Bator F. The anatomy of market failure. Q J Econ. 1958;72(3):351–79.

27. Alexandra Wright, Katherine E. Smith, Mark Hellowell. Policy lessons from health taxes: a systematic review of empirical studies. BMC Public Health. 2017;17(1):1–14.

28. Faulkner G, Grootendorst P, Nguyen VH, Andreyeva T. Economic instruments for obesity prevention: results of a scoping review and modified delphi survey. Int J Behav Nutr Phys Act. 2011;(8):109.

29. Sassi F, Belloni A, Capobianco C. The Role of Fiscal Policies in Health Promotion.

30. Thow AM, Downs S, Jan S. A systematic review of the effectiveness of food taxes and subsidies to improve diets: Understanding the recent evidence. Nutr Rev. 2014;72(9):551–65.

31. Taxation and Sugar-Sweetened Beverages- Position of Dietitians of Canada [Internet]. Dietitians of Canada; 2016 Feb. Available from: https://www.dietitians.ca/Downloads/Public/DC-Position-SSBs-and-taxation.aspx

32. An R. Effectiveness of subsidies in promoting healthy food purchases and consumption: a review of field experiments. Public Health Nutr. 2013;16(7):1215–28.

33. Niebylski M, Redburn K, Duhaney T, Campbell N. Healthy food subsidies and unhealthy food taxation: A systematic review of the evidence. Nutrition. 2015;31:287–95.

34. Morland K, Diez A, Wing S. Supermarkets, other food stores and obesity: The Atherosclerosis risk in communities study. Am J Prev Med. 2006 Apr;30(4):333–9.

35. Ford P, Dzewaltowski D. Disparities in obesity prevalence due to variation in the retail food environment: three testable hypotheses. 2008 Mar 26;66(4).

133

36. Lang B, Harries C, Manon M, Tucker J, Kim E, Smith P. The Healthy Food Financing Handbook: From Advocacy to Implementation [Internet]. Food Trust; 2013. Available from: http://thefoodtrust.org/uploads/media_items/hffhandbookfinal.original.pdf

37. Mah C, Minaker L, Jameson K, Taylor K, Graham M. An introduction to the healthy corner store intervention model. Can J Public Health. 2017;108(3):320–4.

38. Food Trust. Healthy Corner Store Initiative | Overview [Internet]. Food Trust; Available from: http://thefoodtrust.org/uploads/media_items/healthy-corner-store-overview.original.pdf

39. Pinard CA, Byker Shanks C, Harden SM, Yaroch AL. An integrative literature review of small food store research across urban and rural communities in the U.S. Prev Med Rep. 2016;3:324–32.

40. Newfoundland and Labrador. Department of Health and Community Services Strategic Plan 2014-2017 [Internet]. n.d. Available from: http://www.health.gov.nl.ca/health/publications/Department_of_Health_and_Community _Services_Strategic_Plan_2014-17.pdf

41. Newfoundland and Labrador. Eating Healthier in Newfoundland and Labrador: Provincial Food and Nutrition Framework and Action Plan Phase 1: (2006-2008) [Internet]. 2006. Available from: http://www.cssd.gov.nl.ca/healthyliving/pdf/provincialnutritionplan.pdf

42. Newfoundland and Labrador. Department of Health and Community Services (2015). Community Healthy Living Fund [Internet]. Available from: http://www.cssd.gov.nl.ca/grants/chl_fund.html

43. Newfoundland and Labrador. Community Healthy Living Fund 2016-2017 [Internet]. Available from: http://www.cssd.gov.nl.ca/grants/pdf/chl_recipients_16-17.pdf

44. Nutrition North Canada [Internet]. Government of Canada. [cited 2017 Nov 25]. Available from: http://www.nutritionnorthcanada.gc.ca/eng/1415538638170/1415538670874

45. Nutrition North Canada Engagement 2016: Final report of what we heard [Internet]. Interis BDO for Governemnt of Canada; Available from: http://www.nutritionnorthcanada.gc.ca/eng/1491505202346/1491505247821#chp5

46. Dean W, Sharkey J. Rural and urban differences in the associations between characteristics of the community food environment and fruit and vegetable intake. J Nutr Educ Behav. 2011 Nov;43(6):426–33.

134

47. Public Health Agency of Canada, Canadian Institute of Health Information. Obesity in Canada [Internet]. Ottawa Ontario: Public Health Agency of Canada and the Canadian Institute for Health Information; 2011 [cited 2017 Apr 26]. Available from: https://secure.cihi.ca/free_products/Obesity_in_canada_2011_en.pdf

48. Health Canada. Measuring the food environment in Canada [Internet]. Ottawa Ontario; 2013 Oct. Available from: http://publications.gc.ca/collections/collection_2013/sc-hc/H164-155-2012-eng.pdf

49. The Harris Centre, Memorial University. Vital Signs: A province-wide check-up of the quality of life in Newfoundland & Labrador for 2015 [Internet]. Available from: http://www.mun.ca/harriscentre/vitalsigns/VitalSigns_2015-Revised.pdf

50. Treuhaft S, Karpyn A. The grocery gap: Who has access to healthy food and why it matters [Internet]. 2010. Available from: http://thefoodtrust.org/uploads/media_items/grocerygap.original.pdf

51. U.S Department of Health and Human Services. Healthy Food Financing Initiative [Internet]. OffIce of Community Services. Available from: https://www.acf.hhs.gov/ocs/programs/community-economic-development/healthy-food-financing

52. North American Industry Classification System (NAICS) Canada [Internet]. Ottawa, Ontario: Statistics Agency of Canada; 2016. Available from: http://www.statcan.gc.ca/pub/12-501-x/12- 501-x2016001-eng.pdf

53. The National Healthy Corner Stores Network [Internet]. The Food Trust. Available from: http://thefoodtrust.org/what-we-do/administrative/healthy-corner-stores-network

54. Newfoundland and Labrador. Newfoundland and Labrador Nutritious Food Basket 2011 [Internet]. Department of Health and Community Services; Available from: http://www.foodfirstnl.ca/our-resources/newfoundland-and-labrador-nutritious-food-basket

55. Accessibility-Remoteness (A-R) Index Summary Paper [Internet]. Newfoundland and Labrardor Statistics Agency; 2014. Available from: http://www.mae.gov.nl.ca/for/remoteness_index_information.pdf

56. National Collaborating Centre for Healthy Public Policy. A Framework for Analyzing Public Policies: Practical Guide [Internet]. 2012. Available from: http://www.ncchpp.ca/docs/Guide_framework_analyzing_policies_En.pdf

57. Cummins S, Flint E, Matthews SA. New neighborhood grocery store increased awareness of food access but did not alter dietary habits or obesity. Health Aff Proj Hope. 2014;33(2):283.

135

58. Dubowitz T, Ghosh-Dastidar M, Cohen DA, Beckman R, Steiner ED, Hunter GP, et al. Diet And Perceptions Change With Supermarket Introduction In A Food Desert, But Not Because Of Supermarket Use. Health Aff Proj Hope. 2015;34(11):1858.

59. Ghosh-Dastidar M, Hunter G, Collins RL, Zenk SN, Cummins S, Beckman R, et al. Does opening a supermarket in a food desert change the food environment? Health Place. 2017;46:249–56.

60. Rebecca Harris, Mah C, Minaker L. The Nutrition Environment Measures Survey Newfoundland and Labrador: An Exploratory Analysis into the Significance of Assessing the Retail Food Environment Using Healthy Food Only. Unpublished Manuscript. 2017 Dec;

61. Olendzki B, Procter-Gray E, Wedick N, Patil V, Zheng H, Kane K, et al. Disparities in access to healthy and unhealthy foods in central Massachusetts: implications for public health policy. J Am Coll Nutr. 2015;34(2):150–8.

62. Brimblecombe J, Ferguson M, Chatfield MD, Liberato SC, Gunther A, Ball K, et al. Effect of a price discount and consumer education strategy on food and beverage purchases in remote Indigenous Australia: a stepped-wedge randomised controlled trial. Lancet Public Health. 2017;2(2):e82–95.

63. CS & R Associates. Business concept development & feasibility study: North Central Regina community grocery store [Internet]. 2006. Available from: http://www.nccaregina.ca/wp-content/uploads/2011/11/North_Central_Grocery_Study_March_2006.pdf

64. The Healthy Food Financing Initiative: An Innovative Public-Private Partnership Sparking Economic Development and Improving Health [Internet]. 2017 Oct. Available from: http://www.healthyfoodaccess.org/sites/default/files/HFFI%20Brochure_October%202017%20Update.pdf

65. The Food Trust. Philadelphia’s Healthy Corner Store Initiative 2010-2012 [Internet]. 2012. Available from: http://thefoodtrust.org/uploads/media_items/hcsi-y2report-final.original.pdf

66. Magnus A, Moodie M, Ferguson M, Cobiac L, Liberato S, Brimblecombe J. The economic feasibility of price discounts to improve diet in Australian Aboriginal remote communities. Aust NZ J Public Health. 2016;40(1).

67. Regional Development Fund [Internet]. Tourism, Culture, Industry, Innovation. Available from: http://www.tcii.gov.nl.ca/regionaldev/RDF.html

68. Healthy Corner Stores NL [Internet]. Food First NL. Available from: http://www.foodfirstnl.ca/our-projects/healthy-corner-stores-nl

136

69. AFFIRM: Affordable Food in Remote Manitoba [Internet]. Manitoba Government, Healthy Living and Seniors. Available from: http://www.gov.mb.ca/healthyliving/hlp/nutrition/affirm/

70. Key Issues: Community Services (Libraries, Rural Broadband, Rural Healthy, Social Services, Recreation) [Internet]. Municipalities Newfoundland and Labrador. Available from: https://www.municipalnl.ca/issue/community-services/

71. About Us [Internet]. Eastern Health. 2018. Accessed Jan. 2018. Available from: http://www.easternhealth.ca/AboutEH.aspx 72. Dietitians [Internet]. Eastern Health. 2018. Accessed Jan. 2018. Available from: http://www.easternhealth.ca/WebInWeb.aspx?d=3&id=1643&p=1639

137

SECTION III

SUMMARY

138

CHAPTER 8: Summary & Recommendations for Future Research and Policy Directions

139

8.1 INTRODUCTION

Through a series of manuscripts, this dissertation has critically reviewed, analyzed and

discussed the retail food environment in a rural and remote context using the empirical

example of the retail food environment in rural NL. The objectives of this thesis were to

1) identify the methods and features used to describe and define the healthfulness of the

rural food environment, 2) determine the significance and implications of assessing the

consumer food environment using healthy food only (and not unhealthy) and 3)

investigate and analyze fiscal intervention strategies to improve access to nutritious foods

in rural NL. In this concluding chapter, we will review the key learnings from this series

and finish with recommendations for future research and policy direction.

8.2 KEY AREAS OF INVESTIGATION

Rurality

This thesis considered factors in the rural food environment that may be independent and

distinct from those in urban food environments because of the geographic, demographic,

economic, and epidemiological disparities threatening population health in rural and

remote communities. In addition, prevailing disparities including decreased access to

public transportation, higher rates of poverty and unemployment, lower levels of

education, aging populations and increased burden of disease1 are all critical factors to

consider when measuring and intervening in retail food environments in rural

communities. Rural food environments are also particularly significant in the local

140

context with 41% of the NL population residing in rural areas2; a province leading the

nation in obesity and diet related disease.3

Methodological Approaches

In order to support action, it is necessary to produce evidence that is accessible to

knowledge users, guiding them in an appropriate process through which to make change.

As said best by Dr. Leia Minaker “There are thousands of food environment features that

are measurable, but not all measures are equal in terms of their ability to raise awareness

or inform policy priorities.”4 It is for this reason that we dedicated this dissertation to

exploring the ways in which the food environment is measured and reported, in efforts to

identify approaches to describe rural food environments in a format that is transferable to

prioritized policy development and that can lead to more impactful change in rural

communities.

To date, the Nutrition Environment Measures Survey for Stores (NEMS-S) has

consistently served as the proxy gold standard and is among the most widely used

checklist tool in retail food environment assessment.5,6,7,8 NEMS-S has been found to be

a valid and reliable tool and measures the key constructs of the food environment

(availability, quality, price) making it a logical focus for our review.9 Concentrating on

the NEMS-S checklist also allowed us to apply methods identified in the scoping review

in a local context by conducting empirical investigations using the NEMS-NL data

collected through the Healthy Corner Store NL project.

141

Fiscal Interventions

Finally, this thesis began to examine how fiscal policy can be used in relation to retail

food environments and proposed and analyzed two policy options designed to address

some of the fundamental barriers to food access in rural NL. It was our goal to initiate a

critical discussion surrounding emerging retail food environment strategies, with

particular consideration of the suitability of these interventions for rural communities.

As a consequence of their remote location, rural retailers as well as rural consumers face

many economic barriers to accessing nutritious foods. Due to the low-density populations

in rural communities, low profitability is a perceived and objective barrier for rural

businesses owners. The majority of rural communities lack a full-service grocery store

and many rural residents may depend on convenience stores and corner stores for food

shopping.10,11 However, because of the high cost of distribution, reduced economic

power, and the small size and limited capacity of small rural stores, offering fresh healthy

foods can be challenging.12,13,14 Fiscal policies in the form of grants and subsidization can

help alleviate some of these barriers and work to improve availability and affordability of

nutritious foods in rural areas.

8.3 KEY LEARNINGS

A Scoping Review of NEMS-S Research in Rural Communities

Our review of the literature surrounding Nutrition Environment Measures Surveys in

rural communities provided us with a greater understanding of the methods and measures

142

being used to evaluate the rural and remote communities. Among our key findings were

that just over half of articles included for review explicitly defined and operationalized

the variable of rurality and even fewer provided definitions for store types. This was

concerning because of the numerous ways in which a rural community and store types

have been be defined. Without defining these key variables transparently, it makes the

research less accessible and less meaningful to knowledge users who are unable to

interpret the findings with context. We also learned that rural researchers are

predominately using the NEMS to report on healthy food and beverages only (and not

unhealthy) and about 35% of researchers are reporting on availability only (and not

quality and price).

Availability, quality, and price, are all important features in measuring the foodscape,

especially in rural areas where quality15,16,17 and price11,18,17 have been reported as key

barriers to following a healthy diet. Reporting on healthy food only may not capture the

entire consumer experience and is potentially contributing to a gap in intervening in

consumer exposure to non-nutritious foods. We encourage greater transparency in

reporting methodology; and providing explicit definitions and rationales so that findings

are more accessible to knowledge users and can be used to guide future research and

policy direction. Future researchers should consider elaborating upon all features of the

consumer food environment. In order to inform prioritized action and decision making, it

is critical to have a comprehensive understanding of consumer experience. Measuring and

acting on only one element of the food environment, such as availability, may be rendered

143

ineffective if consumers are unable to afford these items or the items available are of poor

quality or unsafe to consume.

Empirical Investigation into Healthy and Unhealthy Food Assessment

Through secondary analysis of data from the HCSNL project, this thesis further

investigated the significance of evaluating the food environment with healthy food

availability only (compared to considering both healthy and unhealthy food). Our results

indicated many stores offered most unhealthy food and beverage while healthy food

availability was less common and therefore was more varied between store types,

demonstrating why it may be appealing to for those with limited recourses to measure

discrepancies in store healthfulness using healthy food availability only.

However, our results also demonstrated that healthy food availability and unhealthy food

availability are highly correlated, suggesting stores who offer a large variety of healthy

foods also offer a high variety of unhealthy foods. These findings corroborate that we

cannot define store healthfulness or consumer experience based on healthy food

availability alone. Further, unhealthy food availability scores were high in both store

types (both supermarkets and convenience stores), with many stores offering most of the

unhealthy items. To date, most retail food environment interventions have been aimed at

increasing the availability of healthy items13,19,20 and it is plausible that the lack of

unhealthy food assessment is contributing to a subsequent gap in intervention strategies

targeted toward reducing the availability of unhealthy items.

144

Policy Analysis- Fiscal Strategies to Improve Food Access in Rural Newfoundland

The objective of the final paper in the manuscript series was to analyze fiscal policy

options to improve the availability and affordability of nutritious foods in rural NL. We

focused our policy analysis on two proposed options 1) Rural Food Retailer Development

Grant and 2) Small Food Retailer Renovation Grants and Distribution Subsidies. This was

a policy option analysis for the NL context supplemented by a critical examination of

relevant scientific and grey literature.

Both proposed options address some of the fundamental challenges to creating healthy

retail food environments in rural communities. Rural food retailer development grants

increase access to healthy foods by removing the geographic barrier to shopping at larger,

full service grocers, who have the capacity to stock a variety of healthy foods at

affordable prices. Small food retailer renovation grants and distribution subsidies increase

the structural capacity of existing rural retailers to stock healthy foods while the subsidy

alleviates the high cost of distribution and allows retailers to offer nutritious foods at

affordable prices.

Although there has been evidence demonstrating the relationship between proximity and

density of grocery stores and improved diet and health outcomes, there have been

minimal evaluations surrounding the effectiveness of improving diet and health outcomes

by introducing a new grocery store into an undeserved area, especially in rural

communities. There is a chance this option could increase the accessibility to unhealthy

foods and potentially lead to an unintended increase in the purchasing and consumption

145

of these items. Further, this option also requires considerable upfront investment and may

pose more economic risk for the province than the renovation grant and subsidization

program.

Small retail renovation grants work with existing infrastructure and existing retailers and

has lower upfront costs. This option has the added benefit of following in the footsteps of

a pilot healthy corner store implemented in rural NL by Memorial University, Food First

NL and Eastern Health. Both healthy corner store programs and subsidization programs

have been successful in increase the availability, consumption and perceptions

surrounding healthy food.

Based on our analysis, we recommended decision makers consider the second option,

small retailer renovation grants and distribution subsidy. This policy will require a multi-

sectoral approach involving the Department of Health and Community Services with

support from Food First NL, Eastern Health, Regional Economic Development and

Service NL, and most importantly rural retailers and rural community members.

8.4 RECOMMENDATIONS FOR FURTURE RESEARCH & POLICY

DIRECTION

Our series of investigations identified several gaps in measuring, reporting and

intervening in the rural retail food environment. The section below will outline a selection

of the gaps that we feel are among the most significant in terms of producing evidence

146

that will work to identify priority areas for action and should be considered for future

research and policy development

Increase Transparency in Defining and Operationalizing Variables

Our review of the Nutrition Environment Measures survey raised concerns surrounding

the transparency and accessibility of retail food environment research. These concerns

stem from the heterogeneity in the methods used define key variables of interest including

rurality and store type.

Defining Rurality

Discrepancy in operationalizing rurality is not exclusive to food environment research;

the use of non-congruent definitions within and across health-related disciplines is

common.21,22,23 Depending on what elements of rural exposure are most relevant to the

variables of interest, definitions of rurality have been based on population size, density,

proximity, contiguous urban areas, commuting trends, economic activity or the social and

cultural aspects associated with rurality.21, 22,23

Fluctuating definitions for rural and remoteness is methodologically problematic as well

as challenging for decision makers, as it as it often creates a barrier to aggregating and

summarizing findings.23,22 Although a standardized definition for rurality would be

advantageous, rurality is multifaceted and how it is measured and defined will depend on

the context and scope of the research. 21,22,23 Rural demography is complex and a single

definition is not able to justifiably capture all aspects.21, 22,23 Dichotomous definitions are

147

limited in their ability to capture the heterogeneity among degrees of rurality and

remoteness which can mask issues at a local level21,22; experts in rural taxonomy in health

research agree that when possible, hierarchical or multilevel definitions should be

applied.21,22,23 Hart, an expert in rural health and rural taxonomy, describes five elements

of an appropriate rural taxonomy: (1) Measures something explicit and meaningful (2)

Replicable (3) Derived from available, high-quality data (4) Quantifiable and not

subjective and (5) Has on the ground validity.23

How rural is defined will have significant implication on findings and subsequent policy

and decision making, and has potential to bias conclusions if applied inappropriately.21,

22,23 Therefore, rural classification should be applied judiciously, based on the context of

the research question21,22,23, and ideally with consultation from geographic

professionals23. Researchers should explicitly indicate how they chose to define rurality

accompanied with a clear rationale so that evidence users can interpret the findings in an

appropriate context.22 When making comparisons across studies, researchers should be

aware of discrepancies in rurality and definition revision over time, and avoid aggregating

data with dissimilar geographic units.22

Defining Store Types

The foundation of much of food environment research has been the investigation into the

disparities in food accessibility within and among differing store types.25,26,27 This

evidence often informs conclusions regarding the healthfulness of certain store types and

is ultimately used to define individual and community level exposure to healthy or

148

unhealthy food environments.28 If or how a store type is classified, can significantly

impact investigation outcomes, evidence interpretation, future research, decision making

and policy development.29

To date, evidence regarding the healthfulness of differing store types has been conflicting

and lack of explicit store classifications have acted as a barrier to the advancement and

refinement of food environment literature.27 An increasing number of retail food outlets

have begun to carry food and beverage, making classifying store types particularly

difficult in rural communities, where the retail presence is often comprised of these non-

traditional or “hybrid” food outlets (e.g. general stores, pharmacies)30,31 Methods of

classification in the literature have included store size, annual sales, number of cash

registers, number of employees, variety of foods offered, owner interpretation and

federally established industry classification codes.10,27,32

Despite efforts within the field33,34, there is yet to be a standardized classification system

recognized as the gold standard. However, the North American Industry Classification

System (NAICS) is among the most widely used classification systems in nutrition

environment literature.10,27,35 A strength of NAICS codes is that they offer refined

definitions for a range of food retailers and facilitates comparability across Canada and

the United States. Although NAICS codes are standardized definitions, they are still

applied with user discretion which can lead to discrepancies within the literature. Some

secondary databases used for store classification will come with pre-assigned NAICS

codes and the accuracy of these classifications have been questioned in previous

149

literature. 36,29 Secondary databases have been shown to misclassify retailers, assign more

than one code or merge NAICS codes, all of which have potential to bias findings in

research relying on these sources alone. 27,36,29

Consumer food environment researchers should continue to work toward establishing

standardized definitions or guidelines for defining retail food outlets that include a wide

variety of stores that may sell food.27,32 Until then, NAICS codes offer recognized,

detailed definitions for classifying most outlets that sell food and beverage and when used

correctly require minimal subjective interpretation. However, to maintain the integrity of

the NAICS codes it is critical to assign them in a standardized way. When using datasets

with preassigned NAICS codes it is recommended that researchers confirm coding either

manually or through predetermined algorithms29, or through other confirmatory methods

such as ground-truthing. If it is necessary to merge NAICS code categories for statistical

or investigative purposes, it should be best practice to provide sufficient detail for any

modifications made so that knowledge users may consider this when interpreting results

and comparing and using findings.

Comprehensive Assessments and Interventions

Quality and Price

Despite the evidence from qualitative and perceived food environment assessment

indicating that quality and price impacts food accessibility for rural dwellers15,17,37,38, 39,

there has been a considerable lack of research reporting on direct field observation of

quality and price in rural retail stores.26,40

150

Quality

Offering and maintaining high quality supplies of fresh food is often a challenge for small

rural stores due to remote location, access to suppliers, limited turnover, and other

infrastructural limitations.12,39,41 Many fresh and frozen wholesalers have minimum

purchasing requirements that are often too large for small rural retailers to manage and

not all offer delivery to remote areas. 12,39,41 Distribution to rural and remote areas can be

costly and lengthy; if not delivered with appropriate refrigeration equipment, produce can

deteriorate before it reaches the store, reducing the shelf life significantly.12 Infrastructure

within small rural stores may also act as a barrier to quality food storage. A deficiency of

adequate refrigeration and freezer equipment due to reduced store size or high cost,

combined with a lack of knowledge regarding appropriate food handling practices by

store employees can also contribute to a shorter shelf life for both fresh and frozen

produce.12,41 Further, the barriers to quality food distribution and storage can contribute to

a cycle of poor supply and demand: Consumers are less likely to purchase produce if the

quality is poor and due to reduce sales, owners perceive a lack of demand and

profitability and are less motivated to increase availability and quality of these items.12

The original NEMS-S survey evaluates the quality of fresh fruits and vegetables only.

However, research conducted six remote communities in Northern Labrador NL indicate

that the quality of fresh and frozen meats may also be of concern in remote

communities.15 Survey results from a report published by Indian and Northern Affairs

Canada in 2002, revealed that 77% of community members classified frozen meat as poor

(41%) or fair (36%) quality.15 Fresh and frozen meat were the most likely items to be

151

classified as poor, while the majority of fresh and frozen fruit and vegetables were

classified as fair or good, indicating that although the quality of produce is perceived as

generally fair in these communities, the quality of meat is perceived to be particularly

poor.15 These findings come from isolated communities serviced by food mail where they

face a higher level of distribution barriers and therefore may not be generalizable to all

rural communities but findings should be considered in future research. The NEMS-NL

researchers have also communicated that during data collection the majority of meat

offered in the rural foodscape was frozen, not fresh, indicating that frozen meats are a

fundamental aspect of the rural consumer experience.

Quality is a significant element of the consumer shopping experience and decision

making process and should not be overlooked during food environment assessment or

intervention, especially in rural and remote communities where shelf life can be reduced.

Researchers may also want to expand quality assessment to include fresh and frozen

meats.

Price

Having healthy options available can be beneficial only if consumers have the power to

make healthy choices. Research has consistently demonstrated that rural areas have

limited geographic access to supermarkets and large grocery stores and have a higher

density of convenience stores and corner stores.10,11,18,37,42,43 Evidence also suggests the

cost of healthy items (when available) including fresh produce, eggs, low-fat/fat free milk

and whole grains, is substantially higher at convenience stores than at supermarkets and

152

grocery stores.11,18,42 Because many larger supermarkets are chain stores, it may be more

feasible for them to offer lower prices than smaller convenience stores or independent,

locally owned stores.41 It can be difficult for small rural stores to offer healthy produce at

an affordable price because unlike larger stores they don’t have the benefit of purchasing

in bulk which can result in higher purchasing costs and remote locations add to increased

distribution expenses.12,41 The lack of retail competition can also negatively affect food

affordability for rural consumers. In many rural communities, there is only one food

retailer and access to transportation is limited, therefore, consumers do not have the

option to shop for lower prices and store owners are less motivated to offer competitive

prices.41

Socioeconomic disparities, higher rates of unemployment and reliance on social

assistance make rural communities even more vulnerable to the burden of increased food

prices and food insecurity.9,10,11. Even more concerning is extremely remote northern

communities who receive food by air or boat, where the barriers and burden of food costs

and the associated implications, are exponentially exaggerated. Food cost contributes to

other food system issues, prominently, household food insecurity. According to the most

recent data, one in eight households in Canada, encompassing over 4 million adults and

children, are food insecure.44

The methods used to measure and evaluate price will influence the data captured and how

it will be reported and interpreted. Among the articles in the scoping review, price was

measured and evaluated in various ways: the price of the healthy option relative (higher

153

or lower) to the price of the regular option, the price of an item relative (higher or lower)

to the median price of that item in other store types or communities and/or the absolute

price or mean price of an item by store type or community.

Analyzing relative price disparities between healthy and unhealthy items within a store is

effective in examining whether the retail food environment offers health-promoting

environmental features and incentivizes consumers to make healthful choices, but it does

not capture overall price, nor does it allow for comparisons among stores or geographic

areas. For example, if a rural store sells a loaf of white bread for $6.00 and a loaf of

whole grain bread for $6.50 and an urban store sells white bread for $3.50 and whole

grain bread for $4.00, both stores would be scored equally and the overall price

discrepancy between the jurisdictions is not captured. In this case, the price of food and

beverages in the rural community could be overlooked as an area for potential

intervention.

Analyzing price relative to other stores or absolute prices allows researchers to capture

price discrepancies among store types and community characteristics, which is

particularly valuable when evaluating stores and or geographic areas which are known to

have higher absolute food costs. Both relative (internal and external) and absolute

measures are valuable and pertinent measures when it comes to evaluating price, so

depending on study objectives, researchers may want to consider using both measures.

154

Measuring and Intervening in the Exposure to Unhealthy Food and Beverage

A consistent theme throughout the manuscript series was the relative paucity of evidence

for measuring, reporting and intervening in consumer exposure to non-nutritious food.

Through our investigation into the significance of measuring the food environment with

healthy foods only, we learned that unhealthy food is widely available in both

supermarkets and convenience stores and is an important feature when characterizing the

healthfulness of food stores.

Despite the widespread exposure to non-nutritious foods, the majority of retail food

environment interventions target healthy food accessibility only.13,19,20 Although many of

these approaches have been proven to be successful in improving the availability,

purchase, consumption and perceptions of healthy foods, evidence of their impact on

obesity and other diet related health outcomes is less prevalent.13,19 This suggests that

targeting healthy food access is necessary, but potentially insufficient if non-nutritious

foods remain highly available.

Unhealthy food taxation is among the few interventions aimed toward reducing the

purchase and consumption of unhealthy foods. However, many have argued this approach

is regressive and disproportionally burdens low income populations.45,46,47,48,49 Further, an

unhealthy food tax does not actually reduce consumer exposure to non-nutritious foods

but rather reduces accessibility. In contrast to a restrictive policy approach to reducing the

availability of unhealthy food, taxation is in theory a more permissive policy approach to

discourage consumers from purchasing unhealthy foods. Restrictive policy approaches

155

targeted toward decreasing the availability of unhealthy foods has been most prevalent in

other organizational food environments, most notably school and sport and recreational

facilities.50,51 These strategies include restricting the type of food that can be serve and

sold to children and adolescents in school lunch programs, canteens and vending

machines.50,51 Similar efforts have expanded to hospital and other health-care settings.52

Despite the evidence of success in other institutions53, these practices have not yet

translated into the broader retail food environment. Within the literature, one of the only

retail based interventions aimed toward reducing the availability of unhealthy foods in

retail food stores was conducted in Nova Scotia, Canada.20 Researchers removed sugar-

sweetened beverages (SSBs) from a pharmacy over a 123-week period to assess the

impact on the overall community sales of carbonated soft drinks.20 Although the removal

of SSBs did lead to a decrease in the overall community sales of soft drinks, the decline

was statistically insignificant.20 However, encouragingly, the removal of sugar sweetened

beverages did not lead to an increase in the sales of these items in surrounding stores,

suggesting removing sugar sweetened beverages from pharmacies could reduce impulse

purchasing of these items.20

The most evident barrier to implementing regulations surrounding the sales of unhealthy

food and beverage in retail environments is the much-anticipated push back from the food

industry due to the perceived threat to their revenue. This threat may be perceived as even

greater in rural communities where operating costs are higher and profitability margins

are much tighter.54 We have witnessed the power of this push back in 2014 when industry

156

lobbyist successfully repealed the proposed limit of soft drink sizes sold at New York

City restaurants and similar food service establishments.55

Although the power of industry may intimidate advocates for healthy public policy, we

can reflect on similar reactions from the tobacco industry when regulations surrounding

the sale, tax and marketing of tobacco products were introduced.56 Researchers, health

practitioners, public health officials and policy makers must learn from the efforts made

when establishing tobacco regulations and continue to display the detrimental long term

health and economic impacts of diets high sugar, salt and fat and the critical role

obesogenic food environments play in contributing to this epidemic.

Measuring and reporting on accessibility of unhealthy food and beverages in the retail

food environment is the first step to earning support and moving toward acting on

limiting the amount and type of non-nutritious food and beverage sold in our

communities. Evidence that is more accessible to knowledge users can help to guide

change. Conducting formative research and working with retailers to implement small

changes such as those demonstrated by Minaker et. al20, are critical to establishing

evidence based practice while holding stakeholders accountable and maintaining policy

interest and support.

157

8.5 CONCLUSION

The overarching aim for this research was to identify and discuss the contextual factors

that are significant to measuring and intervening in the rural retail food environment.

In order to encourage action and influence change, it is most effective to produce

evidence that is accessible to policymakers, retail owners and other key stakeholders,

guiding them in prioritized and informed decision making. To do this, we need to

promote greater transparency in reporting methodology; and providing explicit definitions

and rationales so that findings are more accessible to knowledge users and can be used to

guide future research and policy direction. Reporting on all aspects of the food

environment including availability, quality and price as well as the exposure to nutritious

and non-nutritious foods are critical to capturing barriers and potential areas for

intervention in rural communities.

Improving food access in rural communities will require a comprehensive and multi-

sectoral effort. In NL, existing strategies are broad and relatively abstract. A more direct

and prioritized approach that address the fundamental barriers to food access in rural

communities is essential to make impactful improvements in population health. By

collaborating government, food retailers and communities to make equitable access to

healthy, nutritious food a provincial priority, the province can take the necessary steps

toward improving the health and vitality of rural communities across Newfoundland and

Labrador.

158

REFERENCES

1. Crosby R, Wendel M, Vanderpool R, Casey B. Rural populations and health

determinants, disparities, and solutions. 1st ed. Wiley; 2012.

2. Statistics Canada. Population, urban and rural by province and territory (Newfoundland and Labrador) [Internet]. 2012. Available from: https://www12.statcan.gc.ca/census-recensement/2011/ref/dict/geo042-eng.cfm

3. Statistics Canada. Statistics Canada. Table 105-0508 - Canadian health characteristics, annual estimates, by age group and sex, Canada (excluding territories) and provinces, occasional CANSIM [Internet]. 2017. Available from: http://www5.statcan.gc.ca/cansim/a47#F14

4. Minaker L. Retail food environments in Canada: Maximizing the impact of research, policy and practice. Can J Public Health. 2016 Jun;107(1):5632.

5. Minaker L, Raine K, Wild T, Nykiforuk C, Thompson M, Frank L. Construct validation of 4 food-environment assessment methods: Adapting a multitrait-multimethod matrix approach for environmental measures. Am J Epidemiol. 2014 Feb;179(4):519–28.

6. Glanz K, Johnson L, Yaroch A, Phillips M, Ayala G, Davis E. Measures of retail food store environments and sales: Reviews and implications for healthy eating initiatives. 2016 Apr;48(4):280–8.

7. Moore L, Diez Roux A, Franco M. Measuring availability of healthy foods agreement between directly measured and self-reported data. Am J Epidemiol. 2012 May;175(10):1037–44.

8. Minaker L, Raine K, Wild T, Nykiforuk C, Thompson M, Frank L. Objective food environments and health outcomes. Am J Prev Med. 2013 Sep;45(3):289–96.

9. Glanz K, Sallis J., Saelens BE, Frank L. Nutrition Environment Measures Survey in stores (NEMS-S): Development and evaluation. Am J Prev Med. 2007 Apr;34(4):282–9.

10. Larson N, Story M, Nelson M. Neighborhood environments: Disparities in acess to healthy foods in the U.S. Am J Prev Med. 2008 Jan;36(1).

11. Liese A, Weis K, Pluto D, Smith E, Lawson A. Food store types, availability and costs of food in a rural environment. J Am Diet Assoc. 2007 Nov;107(11):1916–23.

159

12. Providing Fresh Produce in Small Food Stores: Distribution Challenges & Solutions for Healthy Food Retail [Internet]. ChangeLab Solutions; 2014. Available from: http://www.changelabsolutions.org/sites/default/files/Fresh-Produce-Distribution-Small-Food-Stores_FINAL_20140131.pdf

13. Gittelsohn J, Rowan M, Gadhoke P. Interventions in Small Food Stores to Change the Food Environment, Improve Diet, and Reduce Risk of Chronic Disease. Prev Chronic Dis PrevChronic Dis. 2012;9(Generic).

14. Pinard CA, Byker Shanks C, Harden SM, Yaroch AL. An integrative literature review of small food store research across urban and rural communities in the U.S. Prev Med Rep. 2016;3:324–32.

15. Ladouceur L, Hill F, Indian Northern Affairs Canada. Results of the survey on food quality in six isolated communities in Labrador, March 2001 [Internet]. Ottawa Ontario: Government of Canada; 2002. Available from: http://publications.gc.ca/collections/Collection/R2-186-2001E.pdf

16. Byker Shanks C, Ahmed S, Smith T, Houghtaling B, Jenkins M, Margetts M, et al. Availability, Price, and Quality of Fruits and Vegetables in 12 Rural Montana Counties, 2014. Prev Chronic Dis. 2014 Sep;12(E128).

17. Smith C, Morton L. Rural food deserts: low-income perspectives on food access in Minnesota and Iowa. J Nutr Educ Behav. 2009 Jun;41(3):176–87.

18. Jithitikulchai T, Dean W, Sharkey J. Variations in the availability and price of healthier food options by store type and urban-rural setting. 2012 Dec;7(4):381–400.

19. Calancie L, Leeman J, Pitts SJ, Khan L, Fleischhacker S, Evenson K. Nutrition-Related Policy and Environmental Strategies to Prevent Obesity in Rural Communities: A Systematic Review of the Literature, 2002–2013. Prev Chronic Dis. 2015;12(E57).

20. Minaker L, Mackenzie G, Nguyen N, Azagba S, Cook B, Mah C. An evaluation of the impact of a restrictive retail food environment intervention in a rural community pharmacy setting. BMC Public Health. 2016;16.

21. Hall SA, Kaufman JS, Ricketts TC. Defining urban and rural areas in U.S. epidemiologic studies. J Urban Health Bull N Y Acad Med. 2006;83(2):162.

22. Vanderboom C, Madigan E. Federal definitions of rurality and the impact on nursing research. Res Nurs Health. 2007;30(2):162–75.

23. Hart L, Larson E, Lishner D. Rural definitions for health policy and research. m J Public Health. 2005 Jul;95(7):1149–55.

160

24. Hall S, Kaufman J, Ricketts T. Defining urban and rural areas in U.S. epidemiologic studies. 2006 Mar;83(2):162–75.

25. Walker R, Keane C, Burke J. Disparities and access to healthy food in the United States: A review of food deserts literature. Health Place. 2010 Sep;16(5):876–84.

26. Capsi C, Sorensen G, Subramanian SV, Kawachi I. The local food environment and diet: A systematic review. Health Place. 2012;18(5):1172–87.

27. Gamba R, Schucjter J, Rutt C, Seto E. Measuring the food environment and its effects on obesity in the United States: a systematic review of methods and results. 40(3):464–75.

28. Burgoine T, Harrison F. Comparing the accuracy of two secondary food environment data sources in the UK across socio-economic and urban/rural divides. Int J Health Geogr. 2013;12:2.

29. Powell L, Han E, Zenk S, Khan T, Quinn C, Gibbs K, et al. Field validation of secondary commercial data sources on the retail food outlet environment in the U.S. Health Place. 2011 Sep;17(5):1122–31.

30. Bustillos B, Sharkey J, Anding J, McIntosh A. Availability of more healthful food alternatives in traditional, convenience, and nontraditional types of food stores in two rural Texas counties. J Am Diet Assoc. 2009 May;109(5):883–9.

31. Sharkey J. Measuring potential access to food stores and food-service places in rural areas in the U.S. Am J Prev Med. 2009 Apr;36(4):151–5.

32. Fleischhacker S, Evenson K, Sharkey J, Pitts S, Rodriguez D. Validity of secondary retail food outlet data: a systematic review. Am J Prev Med. 2013 Oct;45(4):462–73.

33. Ohri-Vachaspati P, Martinez D, Yedidia MJ, Petlick N. Improving Data Accuracy of Commercial Food Outlet Databases. Am J Health Promot. 2011;26(2):116–22.

34. Lake AA, Burgoine T, Greenhalgh F, Stamp E, Tyrrell R. The foodscape: Classification and field validation of secondary data sources. Health Place. 2010;16(4):666–73.

35. Kelly B, Flood V, Yeatman H. Measuring local food environments: An overview of available methods and measures. Health Place. 2011 Nov;17(6):1284–93.

36. Liese A, Colabianchi N, Lamichane A, Barnes T, Hibbert J, Porter D, et al. Validation of 3 food outlet databases: completeness and geospatial accuracy in rural and urban food environments. Am J Epidemiol. 172(11):1324–33.

161

37. Dean W, Sharkey J. Rural and urban differences in the associations between characteristics of the community food environment and fruit and vegetable intake. J Nutr Educ Behav. 2011 Nov;43(6):426–33.

38. Maley M, Warren B, Devine C. Perceptions of the environment for eating and exercise in a rural community. J Nutr Educ Behav. 2010;42(3):185–91.

39. Office of Nutrition Policy and Promotion, Health Canada. Working with Grocers to Support Healthy Eating [Internet]. Ottawa Ontario: Government of Canada; Available from: https://www.canada.ca/en/health-canada/services/food-nutrition/healthy-eating/nutrition-policy-reports/working-grocers-support-healthy-eating.html

40. Harris R, Minaker L, Mah C. The Nutrition Environment Measures Survey in rural communities: A scoping review. Unpubl Manuscr. 2017;

41. Flournoy R, Treuhaft S. Healthy Food, Healthy Communities: Improving Access and Opportunities Through Food Retailing [Internet]. 2005. Available from: http://www.policylink.org/sites/default/files/HEALTHYFOOD.pdf

42. Vilaro M, Barnett T. The rural food environment: A survey of food price, availability, and quality in a rural Florida community. Food and Public Health. 2013;3(3):111–8.

43. Michimi A, Wimberly M. Associations of supermarket accessibility with obesity and fruit and vegetable consumption in the conterminous United States. Int J Health Geogr. 2010 Oct;9(49).

44. Statistics Canada. Household food insecurity in Canada statistics and graphics [Internet]. 2018 Jan. Available from: https://www.canada.ca/en/health-canada/services/nutrition-science-research/food-security/household-food-security-statistics-2011-2012.html,

45. Powell L. Food prices and obesity: evidence and policy implications for taxes and subsidies. Milkbank Q. 2009 Mar;87(1):229–57.

46. Alexandra Wright, Katherine E. Smith, Mark Hellowell. Policy lessons from health taxes: a systematic review of empirical studies. BMC Public Health. 2017;17(1):1–14.

47. Faulkner G, Grootendorst P, Nguyen VH, Andreyeva T. Economic instruments for obesity prevention: results of a scoping review and modified delphi survey. Int J Behav Nutr Phys Act. 2011;(8):109.

162

48. Thow AM, Downs S, Jan S. A systematic review of the effectiveness of food taxes and subsidies to improve diets: Understanding the recent evidence. Nutr Rev. 2014;72(9):551–65.

49. Taxation and Sugar-Sweetened Beverages- Position of Dietitians of Canada [Internet]. Dietitians of Canada; 2016 Feb. Available from: https://www.dietitians.ca/Downloads/Public/DC-Position-SSBs-and-taxation.aspx

50. Phorson J, The Nutrition Resource Centre. Policies and Guidelines Shaping the School Food Environment: A review of the Literature [Internet]. Available from: http://opha.on.ca/getmedia/f236fa5f-fce2-4beb-a5ea-e2e474a65957/Policies-and-Guidelines-Shaping-the-School-Food-Envt_Lit-Review_NRC_Aug_2015.pdf.aspx

51. Olstad DL, Downs SM, Raine KD, Berry TR, Mccargar LJ. Improving children’s nutrition environments: a survey of adoption and implementation of nutrition guidelines in recreational facilities. BMC Public Health. 2011;11:423.

52. Hospitals with Healthy Food and Beverage Policies [Internet]. Centre for Science in the Public Interest. Available from: https://cspinet.org/sites/default/files/attachment/health-hospitals-examples.pdf

53. Leonard PS. Do School Junk Food Bans Improve Student Health?: Evidence from Canada. Can Public Policy. 2017;43(2):105–19.

54. Flournoy R. Healthy Food, Healthy Communities: Improving Access and Opportunities Through Food Retailing.

55. Gostin L. Banning large sodas is legal and smart [Internet]. 2013. Available from: https://www.cnn.com/2013/03/13/opinion/gostin-soda-ban/index.html

56. Saloojee Y, Dagli E. Tobacco industry tactics for resisting public policy on health. 2000;78(7):902–10.